Identify
Explore
Discover
Discuss
Summarise
Describe
Last, format your objectives into a numbered list. This is because when you write your thesis or dissertation, you will at times need to make reference to a specific research objective; structuring your research objectives in a numbered list will provide a clear way of doing this.
To bring all this together, let’s compare the first research objective in the previous example with the above guidance:
Research Objective:
1. Develop finite element models using explicit dynamics to mimic mallet blows during cup/shell insertion, initially using simplified experimentally validated foam models to represent the acetabulum.
Checking Against Recommended Approach:
Q: Is it specific? A: Yes, it is clear what the student intends to do (produce a finite element model), why they intend to do it (mimic cup/shell blows) and their parameters have been well-defined ( using simplified experimentally validated foam models to represent the acetabulum ).
Q: Is it measurable? A: Yes, it is clear that the research objective will be achieved once the finite element model is complete.
Q: Is it achievable? A: Yes, provided the student has access to a computer lab, modelling software and laboratory data.
Q: Is it relevant? A: Yes, mimicking impacts to a cup/shell is fundamental to the overall aim of understanding how they deform when impacted upon.
Q: Is it timebound? A: Yes, it is possible to create a limited-scope finite element model in a relatively short time, especially if you already have experience in modelling.
Q: Does it start with a verb? A: Yes, it starts with ‘develop’, which makes the intent of the objective immediately clear.
Q: Is it a numbered list? A: Yes, it is the first research objective in a list of eight.
1. making your research aim too broad.
Having a research aim too broad becomes very difficult to achieve. Normally, this occurs when a student develops their research aim before they have a good understanding of what they want to research. Remember that at the end of your project and during your viva defence , you will have to prove that you have achieved your research aims; if they are too broad, this will be an almost impossible task. In the early stages of your research project, your priority should be to narrow your study to a specific area. A good way to do this is to take the time to study existing literature, question their current approaches, findings and limitations, and consider whether there are any recurring gaps that could be investigated .
Note: Achieving a set of aims does not necessarily mean proving or disproving a theory or hypothesis, even if your research aim was to, but having done enough work to provide a useful and original insight into the principles that underlie your research aim.
Be realistic about what you can achieve in the time you have available. It is natural to want to set ambitious research objectives that require sophisticated data collection and analysis, but only completing this with six months before the end of your PhD registration period is not a worthwhile trade-off.
Each research objective should have its own purpose and distinct measurable outcome. To this effect, a common mistake is to form research objectives which have large amounts of overlap. This makes it difficult to determine when an objective is truly complete, and also presents challenges in estimating the duration of objectives when creating your project timeline. It also makes it difficult to structure your thesis into unique chapters, making it more challenging for you to write and for your audience to read.
Fortunately, this oversight can be easily avoided by using SMART objectives.
Hopefully, you now have a good idea of how to create an effective set of aims and objectives for your research project, whether it be a thesis, dissertation or research paper. While it may be tempting to dive directly into your research, spending time on getting your aims and objectives right will give your research clear direction. This won’t only reduce the likelihood of problems arising later down the line, but will also lead to a more thorough and coherent research project.
Finding a PhD has never been this easy – search for a PhD by keyword, location or academic area of interest.
Join thousands of students.
Join thousands of other students and stay up to date with the latest PhD programmes, funding opportunities and advice.
Writing a research paper, thesis, or dissertation ? If so, you’ll want to state your research objectives in the introduction of your paper to make it clear to your readers what you’re trying to accomplish. But how do you write effective research objectives? In this post, we’ll look at two key topics to help you do this:
For more advice on how to write strong research objectives, see below.
There is an important difference between research aims and research objectives:
For instance, an example research aim could be:
This study will investigate the link between dehydration and the incidence of urinary tract infections (UTIs) in intensive care patients in Australia.
To develop a set of research objectives, you would then break down the various steps involved in meeting said aim. For example:
This study will investigate the link between dehydration and the incidence of urinary tract infections (UTIs) in intensive care patients in Australia. To achieve this, the study objectives w ill include:
Subscribe to our newsletter and get writing tips from our editors straight to your inbox.
Note that the objectives don’t go into any great detail here. The key is to briefly summarize each component of your study. You can save details for how you will conduct the research for the methodology section of your paper.
A great way to refine your research objectives is to use SMART criteria . Borrowed from the world of project management, there are many versions of this system. However, we’re going to focus on developing specific, measurable, achievable, relevant, and timebound objectives.
In other words, a good research objective should be all of the following:
If you follow this system, your research objectives will be much stronger.
Whatever your research aims and objectives, make sure to have your academic writing proofread by the experts!
Our academic editors can help you with research papers and proposals , as well as any other scholarly document you need checking. And this will help to ensure that your academic writing is always clear, concise, and precise.
Submit a free sample document today to trial our services and find out more.
Post A New Comment
5-minute read
Promoting a brand means sharing valuable insights to connect more deeply with your audience, and...
6-minute read
If you’re seeking funding to support your charitable endeavors as a nonprofit organization, you’ll need...
9-minute read
Is your content getting noticed? Capturing and maintaining an audience’s attention is a challenge when...
8-minute read
Are you looking to enhance engagement and captivate your audience through your professional documents? Interactive...
7-minute read
Voice search optimization is rapidly shaping the digital landscape, requiring content professionals to adapt their...
4-minute read
Are you a creative freelancer looking to make a lasting impression on potential clients or...
Formulating research aim and objectives in an appropriate manner is one of the most important aspects of your thesis. This is because research aim and objectives determine the scope, depth and the overall direction of the research. Research question is the central question of the study that has to be answered on the basis of research findings.
Research aim emphasizes what needs to be achieved within the scope of the research, by the end of the research process. Achievement of research aim provides answer to the research question.
Research objectives divide research aim into several parts and address each part separately. Research aim specifies WHAT needs to be studied and research objectives comprise a number of steps that address HOW research aim will be achieved.
As a rule of dumb, there would be one research aim and several research objectives. Achievement of each research objective will lead to the achievement of the research aim.
Consider the following as an example:
Research title: Effects of organizational culture on business profitability: a case study of Virgin Atlantic
Research aim: To assess the effects of Virgin Atlantic organizational culture on business profitability
Following research objectives would facilitate the achievement of this aim:
Figure below illustrates additional examples in formulating research aims and objectives:
Formulation of research question, aim and objectives
Common mistakes in the formulation of research aim relate to the following:
1. Choosing the topic too broadly . This is the most common mistake. For example, a research title of “an analysis of leadership practices” can be classified as too broad because the title fails to answer the following questions:
a) Which aspects of leadership practices? Leadership has many aspects such as employee motivation, ethical behaviour, strategic planning, change management etc. An attempt to cover all of these aspects of organizational leadership within a single research will result in an unfocused and poor work.
b) An analysis of leadership practices in which country? Leadership practices tend to be different in various countries due to cross-cultural differences, legislations and a range of other region-specific factors. Therefore, a study of leadership practices needs to be country-specific.
c) Analysis of leadership practices in which company or industry? Similar to the point above, analysis of leadership practices needs to take into account industry-specific and/or company-specific differences, and there is no way to conduct a leadership research that relates to all industries and organizations in an equal manner.
Accordingly, as an example “a study into the impacts of ethical behaviour of a leader on the level of employee motivation in US healthcare sector” would be a more appropriate title than simply “An analysis of leadership practices”.
2. Setting an unrealistic aim . Formulation of a research aim that involves in-depth interviews with Apple strategic level management by an undergraduate level student can be specified as a bit over-ambitious. This is because securing an interview with Apple CEO Tim Cook or members of Apple Board of Directors might not be easy. This is an extreme example of course, but you got the idea. Instead, you may aim to interview the manager of your local Apple store and adopt a more feasible strategy to get your dissertation completed.
3. Choosing research methods incompatible with the timeframe available . Conducting interviews with 20 sample group members and collecting primary data through 2 focus groups when only three months left until submission of your dissertation can be very difficult, if not impossible. Accordingly, timeframe available need to be taken into account when formulating research aims and objectives and selecting research methods.
Moreover, research objectives need to be formulated according to SMART principle,
where the abbreviation stands for specific, measurable, achievable, realistic, and time-bound.
Study employee motivation of Coca-Cola | To study the impacts of management practices on the levels of employee motivation at Coca-Cola US by December 5, 2022
|
Analyze consumer behaviour in catering industry
| Analyzing changes in consumer behaviour in catering industry in the 21 century in the UK by March 1, 2022 |
Recommend Toyota Motor Corporation management on new market entry strategy
| Formulating recommendations to Toyota Motor Corporation management on the choice of appropriate strategy to enter Vietnam market by June 9, 2022
|
Analyze the impact of social media marketing on business
| Assessing impacts of integration of social media into marketing strategy on the level of brand awareness by March 30, 2022
|
Finding out about time management principles used by Accenture managers | Identifying main time-management strategies used by managers of Accenture France by December 1, 2022 |
Examples of SMART research objectives
At the conclusion part of your research project you will need to reflect on the level of achievement of research aims and objectives. In case your research aims and objectives are not fully achieved by the end of the study, you will need to discuss the reasons. These may include initial inappropriate formulation of research aims and objectives, effects of other variables that were not considered at the beginning of the research or changes in some circumstances during the research process.
John Dudovskiy
Home » Research Topics – Ideas and Examples
Table of Contents
Definition:
Research topic is a specific subject or area of interest that a researcher wants to investigate or explore in-depth through research. It is the overarching theme or question that guides a research project and helps to focus the research activities towards a clear objective.
You can Choose a Research Topic by following the below guide:
One of the most important factors to consider when choosing a research topic is your personal interest. This is because you will be spending a considerable amount of time researching and writing about the topic, so it’s essential that you are genuinely interested and passionate about it. Start by brainstorming a list of potential research topics based on your interests, hobbies, or areas of expertise. You can also consider the courses that you’ve enjoyed the most or the topics that have stood out to you in your readings.
Before deciding on a research topic, you need to understand what has already been written about it. Conducting a preliminary review of the existing literature in your field can help you identify gaps in knowledge, inconsistencies in findings, or unanswered questions that you can explore further. You can do this by reading academic articles, books, and other relevant sources in your field. Make notes of the themes or topics that emerge and use this information to guide your research question.
Your academic advisor or a mentor in your field can provide you with valuable insights and guidance on choosing a research topic. They can help you identify areas of interest, suggest potential research questions, and provide feedback on the feasibility of your research proposal. They can also direct you towards relevant literature and resources that can help you develop your research further.
The research topic you choose should be manageable within the time and resource constraints of your project. Be mindful of the scope of your research and ensure that you are not trying to tackle a topic that is too broad or too narrow. If your topic is too broad, you may find it challenging to conduct a comprehensive analysis, while if it’s too narrow, you may struggle to find enough material to support your research.
Discussing potential research topics with your peers or colleagues can help you generate new ideas and perspectives. They may have insights or expertise that you haven’t considered, and their feedback can help you refine your research question. You can also join academic groups or attend conferences in your field to network with other researchers and get inspiration for your research.
Choose a research topic that is relevant to your field of study and has the potential to contribute to the existing knowledge. You can consider the latest trends and emerging issues in your field to identify topics that are both relevant and interesting. Conducting research on a topic that is timely and relevant can also increase the likelihood of getting published or presenting your research at conferences.
While it’s essential to choose a research topic that aligns with your interests and expertise, you should also be open to exploring new ideas or topics that may be outside of your comfort zone. Consider researching a topic that challenges your assumptions or introduces new perspectives that you haven’t considered before. You may discover new insights or perspectives that can enrich your research and contribute to your growth as a researcher.
A research topic typically consists of several components that help to define and clarify the subject matter of the research project. These components include:
The purpose of a research topic is to identify a specific area of inquiry that the researcher wants to explore and investigate. A research topic is typically a broad area of interest that requires further exploration and refinement through the research process. It provides a clear focus and direction for the research project, and helps to define the research questions and objectives. A well-defined research topic also helps to ensure that the research is relevant and useful, and can contribute to the existing body of knowledge in the field. Ultimately, the purpose of a research topic is to generate new insights, knowledge, and understanding about a particular phenomenon, issue, or problem.
some common characteristics of a well-defined research topic include:
Here are some examples of research topics that are currently relevant and in-demand in various fields:
Here are some Research Topics Ideas from different fields:
Researcher, Academic Writer, Web developer
25+ Practical Examples & Ideas To Help You Get Started
By: Derek Jansen (MBA) | October 2023
A well-crafted research question (or set of questions) sets the stage for a robust study and meaningful insights. But, if you’re new to research, it’s not always clear what exactly constitutes a good research question. In this post, we’ll provide you with clear examples of quality research questions across various disciplines, so that you can approach your research project with confidence!
Let’s start by looking at some examples of research questions that you might encounter within the discipline of psychology.
How does sleep quality affect academic performance in university students?
This question is specific to a population (university students) and looks at a direct relationship between sleep and academic performance, both of which are quantifiable and measurable variables.
What factors contribute to the onset of anxiety disorders in adolescents?
The question narrows down the age group and focuses on identifying multiple contributing factors. There are various ways in which it could be approached from a methodological standpoint, including both qualitatively and quantitatively.
Do mindfulness techniques improve emotional well-being?
This is a focused research question aiming to evaluate the effectiveness of a specific intervention.
How does early childhood trauma impact adult relationships?
This research question targets a clear cause-and-effect relationship over a long timescale, making it focused but comprehensive.
Is there a correlation between screen time and depression in teenagers?
This research question focuses on an in-demand current issue and a specific demographic, allowing for a focused investigation. The key variables are clearly stated within the question and can be measured and analysed (i.e., high feasibility).
Next, let’s look at some examples of well-articulated research questions within the business and management realm.
How do leadership styles impact employee retention?
This is an example of a strong research question because it directly looks at the effect of one variable (leadership styles) on another (employee retention), allowing from a strongly aligned methodological approach.
What role does corporate social responsibility play in consumer choice?
Current and precise, this research question can reveal how social concerns are influencing buying behaviour by way of a qualitative exploration.
Does remote work increase or decrease productivity in tech companies?
Focused on a particular industry and a hot topic, this research question could yield timely, actionable insights that would have high practical value in the real world.
How do economic downturns affect small businesses in the homebuilding industry?
Vital for policy-making, this highly specific research question aims to uncover the challenges faced by small businesses within a certain industry.
Which employee benefits have the greatest impact on job satisfaction?
By being straightforward and specific, answering this research question could provide tangible insights to employers.
Next, let’s look at some potential research questions within the education, training and development domain.
How does class size affect students’ academic performance in primary schools?
This example research question targets two clearly defined variables, which can be measured and analysed relatively easily.
Do online courses result in better retention of material than traditional courses?
Timely, specific and focused, answering this research question can help inform educational policy and personal choices about learning formats.
What impact do US public school lunches have on student health?
Targeting a specific, well-defined context, the research could lead to direct changes in public health policies.
To what degree does parental involvement improve academic outcomes in secondary education in the Midwest?
This research question focuses on a specific context (secondary education in the Midwest) and has clearly defined constructs.
What are the negative effects of standardised tests on student learning within Oklahoma primary schools?
This research question has a clear focus (negative outcomes) and is narrowed into a very specific context.
Shifting to a different field, let’s look at some examples of research questions within the healthcare space.
What are the most effective treatments for chronic back pain amongst UK senior males?
Specific and solution-oriented, this research question focuses on clear variables and a well-defined context (senior males within the UK).
How do different healthcare policies affect patient satisfaction in public hospitals in South Africa?
This question is has clearly defined variables and is narrowly focused in terms of context.
Which factors contribute to obesity rates in urban areas within California?
This question is focused yet broad, aiming to reveal several contributing factors for targeted interventions.
Does telemedicine provide the same perceived quality of care as in-person visits for diabetes patients?
Ideal for a qualitative study, this research question explores a single construct (perceived quality of care) within a well-defined sample (diabetes patients).
Which lifestyle factors have the greatest affect on the risk of heart disease?
This research question aims to uncover modifiable factors, offering preventive health recommendations.
Last but certainly not least, let’s look at a few examples of research questions within the computer science world.
What are the perceived risks of cloud-based storage systems?
Highly relevant in our digital age, this research question would align well with a qualitative interview approach to better understand what users feel the key risks of cloud storage are.
Which factors affect the energy efficiency of data centres in Ohio?
With a clear focus, this research question lays a firm foundation for a quantitative study.
How do TikTok algorithms impact user behaviour amongst new graduates?
While this research question is more open-ended, it could form the basis for a qualitative investigation.
What are the perceived risk and benefits of open-source software software within the web design industry?
Practical and straightforward, the results could guide both developers and end-users in their choices.
In this post, we’ve tried to provide a wide range of research question examples to help you get a feel for what research questions look like in practice. That said, it’s important to remember that these are just examples and don’t necessarily equate to good research topics . If you’re still trying to find a topic, check out our topic megalist for inspiration.
This post was based on one of our popular Research Bootcamps . If you're working on a research project, you'll definitely want to check this out ...
Research ideas on Political Science
Your email address will not be published. Required fields are marked *
Save my name, email, and website in this browser for the next time I comment.
Ai generator.
Research objectives are specific goals or purposes that guide a study or investigation. They are clearly defined statements that outline what the researcher aims to achieve through their research . These objectives help to focus the study, provide direction, and establish the scope of the research design . They typically include the main questions or problems the research seeks to address and are essential for designing the methodology, data collection, and analysis processes. By defining research objectives , researchers can ensure their study remains on track and addresses the key issues relevant to their topic.
Research objectives are clear, specific goals that guide a study’s direction and scope. They outline what the researcher aims to achieve, helping to focus the research, design methodologies, and guide data collection and analysis. These objectives ensure the research stays on track and addresses key issues relevant to the topic.
Research objectives are crucial because they provide clear direction and focus for a study, ensuring that the research stays on track and addresses the specific goals set by the researcher. They help in the formulation of research questions and the design of the methodology, guiding data collection and analysis processes. Well-defined objectives make it easier to measure the study’s success and ensure that the findings are relevant and meaningful. They also enhance the credibility and reliability of the research by outlining a precise plan, making it easier for others to understand and replicate the study.
Descriptive objectives aim to describe the characteristics or functions of a particular phenomenon or population. These objectives focus on answering the “what” aspect of research. Example : To describe the demographic characteristics of smartphone users in the United States.
Exploratory objectives aim to explore new areas where little information is available. They seek to gain insights and familiarize the researcher with the subject area. Example : To explore the potential factors influencing consumer preferences for electric vehicles.
Explanatory objectives aim to explain the relationships or causality between variables. These objectives focus on understanding the “why” and “how” aspects of research. Example : To explain the relationship between social media usage and academic performance among college students.
Predictive objectives aim to predict the future trends, behaviors, or outcomes based on current data or trends. These objectives are used to forecast and anticipate future scenarios. Example : To predict the impact of climate change on agricultural productivity over the next decade.
Evaluative objectives aim to assess the effectiveness or impact of an intervention, program, or policy. These objectives focus on determining the success or value of something. Example : To evaluate the effectiveness of a new employee training program on job performance.
Research objectives are crucial components of any study as they define the purpose and goals of the research. Well-crafted research objectives provide clarity, direction, and focus to the study. Here are the key characteristics of research objectives:
Research objectives should be clear and precise, leaving no room for ambiguity. They should clearly state what the research intends to achieve. Example: Specific Objective: “To determine the impact of social media marketing on consumer purchasing decisions.”
Objectives should be quantifiable, allowing researchers to assess the extent to which they have been achieved. This involves using metrics or indicators that can be measured. Example: Measurable Objective: “To measure the increase in sales by 15% after implementing a social media marketing campaign.”
The objectives should be realistic and attainable within the scope and resources of the study. Setting achievable goals ensures that the research can be completed successfully. Example: Achievable Objective: “To survey 500 consumers within a three-month period to understand their social media usage patterns.”
Objectives must be relevant to the research problem and aligned with the overall purpose of the study. They should address the key issues and contribute to solving the research problem. Example: Relevant Objective: “To analyze the relationship between social media engagement and brand loyalty among teenagers.”
Objectives should have a clear timeframe within which they are to be achieved. This helps in planning and maintaining the research schedule. Example: Time-bound Objective: “To complete data collection within six months and publish findings within one year.”
1. identify the research problem.
Advantages of research objectives.
Why are research objectives important.
Research objectives provide direction and focus for the study, ensuring that the research stays on track and addresses relevant questions.
Formulate research objectives by identifying key questions your research aims to answer, ensuring they are specific, measurable, achievable, relevant, and time-bound (SMART).
Research objectives outline the goals of the study, while research questions specify what the researcher aims to find out.
Yes, research objectives can be refined or adjusted as the study progresses, especially if new insights emerge.
The number of research objectives depends on the scope of the study but typically ranges from two to five.
Research objectives guide the study, while hypotheses are testable predictions derived from these objectives.
Yes, research objectives can be either qualitative, focusing on understanding phenomena, or quantitative, focusing on measuring variables.
Prioritize research objectives based on their relevance to the research problem and feasibility within the study’s constraints.
Research objectives help structure the literature review, guiding the selection of relevant studies and identifying gaps in existing research.
Research objectives determine the type of data needed and the appropriate methods for collecting this data.
Text prompt
10 Examples of Public speaking
20 Examples of Gas lighting
An official website of the United States government
The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.
The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.
Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .
Patricia farrugia.
* Michael G. DeGroote School of Medicine, the
† Division of Orthopaedic Surgery and the
‡ Departments of Surgery and
§ Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, Ont
There is an increasing familiarity with the principles of evidence-based medicine in the surgical community. As surgeons become more aware of the hierarchy of evidence, grades of recommendations and the principles of critical appraisal, they develop an increasing familiarity with research design. Surgeons and clinicians are looking more and more to the literature and clinical trials to guide their practice; as such, it is becoming a responsibility of the clinical research community to attempt to answer questions that are not only well thought out but also clinically relevant. The development of the research question, including a supportive hypothesis and objectives, is a necessary key step in producing clinically relevant results to be used in evidence-based practice. A well-defined and specific research question is more likely to help guide us in making decisions about study design and population and subsequently what data will be collected and analyzed. 1
In this article, we discuss important considerations in the development of a research question and hypothesis and in defining objectives for research. By the end of this article, the reader will be able to appreciate the significance of constructing a good research question and developing hypotheses and research objectives for the successful design of a research study. The following article is divided into 3 sections: research question, research hypothesis and research objectives.
Interest in a particular topic usually begins the research process, but it is the familiarity with the subject that helps define an appropriate research question for a study. 1 Questions then arise out of a perceived knowledge deficit within a subject area or field of study. 2 Indeed, Haynes suggests that it is important to know “where the boundary between current knowledge and ignorance lies.” 1 The challenge in developing an appropriate research question is in determining which clinical uncertainties could or should be studied and also rationalizing the need for their investigation.
Increasing one’s knowledge about the subject of interest can be accomplished in many ways. Appropriate methods include systematically searching the literature, in-depth interviews and focus groups with patients (and proxies) and interviews with experts in the field. In addition, awareness of current trends and technological advances can assist with the development of research questions. 2 It is imperative to understand what has been studied about a topic to date in order to further the knowledge that has been previously gathered on a topic. Indeed, some granting institutions (e.g., Canadian Institute for Health Research) encourage applicants to conduct a systematic review of the available evidence if a recent review does not already exist and preferably a pilot or feasibility study before applying for a grant for a full trial.
In-depth knowledge about a subject may generate a number of questions. It then becomes necessary to ask whether these questions can be answered through one study or if more than one study needed. 1 Additional research questions can be developed, but several basic principles should be taken into consideration. 1 All questions, primary and secondary, should be developed at the beginning and planning stages of a study. Any additional questions should never compromise the primary question because it is the primary research question that forms the basis of the hypothesis and study objectives. It must be kept in mind that within the scope of one study, the presence of a number of research questions will affect and potentially increase the complexity of both the study design and subsequent statistical analyses, not to mention the actual feasibility of answering every question. 1 A sensible strategy is to establish a single primary research question around which to focus the study plan. 3 In a study, the primary research question should be clearly stated at the end of the introduction of the grant proposal, and it usually specifies the population to be studied, the intervention to be implemented and other circumstantial factors. 4
Hulley and colleagues 2 have suggested the use of the FINER criteria in the development of a good research question ( Box 1 ). The FINER criteria highlight useful points that may increase the chances of developing a successful research project. A good research question should specify the population of interest, be of interest to the scientific community and potentially to the public, have clinical relevance and further current knowledge in the field (and of course be compliant with the standards of ethical boards and national research standards).
Feasible | ||
Interesting | ||
Novel | ||
Ethical | ||
Relevant |
Adapted with permission from Wolters Kluwer Health. 2
Whereas the FINER criteria outline the important aspects of the question in general, a useful format to use in the development of a specific research question is the PICO format — consider the population (P) of interest, the intervention (I) being studied, the comparison (C) group (or to what is the intervention being compared) and the outcome of interest (O). 3 , 5 , 6 Often timing (T) is added to PICO ( Box 2 ) — that is, “Over what time frame will the study take place?” 1 The PICOT approach helps generate a question that aids in constructing the framework of the study and subsequently in protocol development by alluding to the inclusion and exclusion criteria and identifying the groups of patients to be included. Knowing the specific population of interest, intervention (and comparator) and outcome of interest may also help the researcher identify an appropriate outcome measurement tool. 7 The more defined the population of interest, and thus the more stringent the inclusion and exclusion criteria, the greater the effect on the interpretation and subsequent applicability and generalizability of the research findings. 1 , 2 A restricted study population (and exclusion criteria) may limit bias and increase the internal validity of the study; however, this approach will limit external validity of the study and, thus, the generalizability of the findings to the practical clinical setting. Conversely, a broadly defined study population and inclusion criteria may be representative of practical clinical practice but may increase bias and reduce the internal validity of the study.
Population (patients) | ||
Intervention (for intervention studies only) | ||
Comparison group | ||
Outcome of interest | ||
Time |
A poorly devised research question may affect the choice of study design, potentially lead to futile situations and, thus, hamper the chance of determining anything of clinical significance, which will then affect the potential for publication. Without devoting appropriate resources to developing the research question, the quality of the study and subsequent results may be compromised. During the initial stages of any research study, it is therefore imperative to formulate a research question that is both clinically relevant and answerable.
The primary research question should be driven by the hypothesis rather than the data. 1 , 2 That is, the research question and hypothesis should be developed before the start of the study. This sounds intuitive; however, if we take, for example, a database of information, it is potentially possible to perform multiple statistical comparisons of groups within the database to find a statistically significant association. This could then lead one to work backward from the data and develop the “question.” This is counterintuitive to the process because the question is asked specifically to then find the answer, thus collecting data along the way (i.e., in a prospective manner). Multiple statistical testing of associations from data previously collected could potentially lead to spuriously positive findings of association through chance alone. 2 Therefore, a good hypothesis must be based on a good research question at the start of a trial and, indeed, drive data collection for the study.
The research or clinical hypothesis is developed from the research question and then the main elements of the study — sampling strategy, intervention (if applicable), comparison and outcome variables — are summarized in a form that establishes the basis for testing, statistical and ultimately clinical significance. 3 For example, in a research study comparing computer-assisted acetabular component insertion versus freehand acetabular component placement in patients in need of total hip arthroplasty, the experimental group would be computer-assisted insertion and the control/conventional group would be free-hand placement. The investigative team would first state a research hypothesis. This could be expressed as a single outcome (e.g., computer-assisted acetabular component placement leads to improved functional outcome) or potentially as a complex/composite outcome; that is, more than one outcome (e.g., computer-assisted acetabular component placement leads to both improved radiographic cup placement and improved functional outcome).
However, when formally testing statistical significance, the hypothesis should be stated as a “null” hypothesis. 2 The purpose of hypothesis testing is to make an inference about the population of interest on the basis of a random sample taken from that population. The null hypothesis for the preceding research hypothesis then would be that there is no difference in mean functional outcome between the computer-assisted insertion and free-hand placement techniques. After forming the null hypothesis, the researchers would form an alternate hypothesis stating the nature of the difference, if it should appear. The alternate hypothesis would be that there is a difference in mean functional outcome between these techniques. At the end of the study, the null hypothesis is then tested statistically. If the findings of the study are not statistically significant (i.e., there is no difference in functional outcome between the groups in a statistical sense), we cannot reject the null hypothesis, whereas if the findings were significant, we can reject the null hypothesis and accept the alternate hypothesis (i.e., there is a difference in mean functional outcome between the study groups), errors in testing notwithstanding. In other words, hypothesis testing confirms or refutes the statement that the observed findings did not occur by chance alone but rather occurred because there was a true difference in outcomes between these surgical procedures. The concept of statistical hypothesis testing is complex, and the details are beyond the scope of this article.
Another important concept inherent in hypothesis testing is whether the hypotheses will be 1-sided or 2-sided. A 2-sided hypothesis states that there is a difference between the experimental group and the control group, but it does not specify in advance the expected direction of the difference. For example, we asked whether there is there an improvement in outcomes with computer-assisted surgery or whether the outcomes worse with computer-assisted surgery. We presented a 2-sided test in the above example because we did not specify the direction of the difference. A 1-sided hypothesis states a specific direction (e.g., there is an improvement in outcomes with computer-assisted surgery). A 2-sided hypothesis should be used unless there is a good justification for using a 1-sided hypothesis. As Bland and Atlman 8 stated, “One-sided hypothesis testing should never be used as a device to make a conventionally nonsignificant difference significant.”
The research hypothesis should be stated at the beginning of the study to guide the objectives for research. Whereas the investigators may state the hypothesis as being 1-sided (there is an improvement with treatment), the study and investigators must adhere to the concept of clinical equipoise. According to this principle, a clinical (or surgical) trial is ethical only if the expert community is uncertain about the relative therapeutic merits of the experimental and control groups being evaluated. 9 It means there must exist an honest and professional disagreement among expert clinicians about the preferred treatment. 9
Designing a research hypothesis is supported by a good research question and will influence the type of research design for the study. Acting on the principles of appropriate hypothesis development, the study can then confidently proceed to the development of the research objective.
The primary objective should be coupled with the hypothesis of the study. Study objectives define the specific aims of the study and should be clearly stated in the introduction of the research protocol. 7 From our previous example and using the investigative hypothesis that there is a difference in functional outcomes between computer-assisted acetabular component placement and free-hand placement, the primary objective can be stated as follows: this study will compare the functional outcomes of computer-assisted acetabular component insertion versus free-hand placement in patients undergoing total hip arthroplasty. Note that the study objective is an active statement about how the study is going to answer the specific research question. Objectives can (and often do) state exactly which outcome measures are going to be used within their statements. They are important because they not only help guide the development of the protocol and design of study but also play a role in sample size calculations and determining the power of the study. 7 These concepts will be discussed in other articles in this series.
From the surgeon’s point of view, it is important for the study objectives to be focused on outcomes that are important to patients and clinically relevant. For example, the most methodologically sound randomized controlled trial comparing 2 techniques of distal radial fixation would have little or no clinical impact if the primary objective was to determine the effect of treatment A as compared to treatment B on intraoperative fluoroscopy time. However, if the objective was to determine the effect of treatment A as compared to treatment B on patient functional outcome at 1 year, this would have a much more significant impact on clinical decision-making. Second, more meaningful surgeon–patient discussions could ensue, incorporating patient values and preferences with the results from this study. 6 , 7 It is the precise objective and what the investigator is trying to measure that is of clinical relevance in the practical setting.
The following is an example from the literature about the relation between the research question, hypothesis and study objectives:
Study: Warden SJ, Metcalf BR, Kiss ZS, et al. Low-intensity pulsed ultrasound for chronic patellar tendinopathy: a randomized, double-blind, placebo-controlled trial. Rheumatology 2008;47:467–71.
Research question: How does low-intensity pulsed ultrasound (LIPUS) compare with a placebo device in managing the symptoms of skeletally mature patients with patellar tendinopathy?
Research hypothesis: Pain levels are reduced in patients who receive daily active-LIPUS (treatment) for 12 weeks compared with individuals who receive inactive-LIPUS (placebo).
Objective: To investigate the clinical efficacy of LIPUS in the management of patellar tendinopathy symptoms.
The development of the research question is the most important aspect of a research project. A research project can fail if the objectives and hypothesis are poorly focused and underdeveloped. Useful tips for surgical researchers are provided in Box 3 . Designing and developing an appropriate and relevant research question, hypothesis and objectives can be a difficult task. The critical appraisal of the research question used in a study is vital to the application of the findings to clinical practice. Focusing resources, time and dedication to these 3 very important tasks will help to guide a successful research project, influence interpretation of the results and affect future publication efforts.
FINER = feasible, interesting, novel, ethical, relevant; PICOT = population (patients), intervention (for intervention studies only), comparison group, outcome of interest, time.
Competing interests: No funding was received in preparation of this paper. Dr. Bhandari was funded, in part, by a Canada Research Chair, McMaster University.
Integrations
What's new?
In-Product Prompts
Participant Management
Interview Studies
Prototype Testing
Card Sorting
Tree Testing
Live Website Testing
Automated Reports
Templates Gallery
Choose from our library of pre-built mazes to copy, customize, and share with your own users
Browse all templates
Financial Services
Tech & Software
Product Designers
Product Managers
User Researchers
By use case
Concept & Idea Validation
Wireframe & Usability Test
Content & Copy Testing
Feedback & Satisfaction
Content Hub
Educational resources for product, research and design teams
Explore all resources
Question Bank
Research Maturity Model
Guides & Reports
Help Center
Future of User Research Report
The Optimal Path Podcast
User Research
Aug 19, 2024 • 17 minutes read
Qualitative research uncovers in-depth user insights, but what does it look like? Here are seven methods and examples to help you get the data you need.
Armin Tanovic
Behind every what, there’s a why . Qualitative research is how you uncover that why. It enables you to connect with users and understand their thoughts, feelings, wants, needs, and pain points.
There’s many methods for conducting qualitative research, and many objectives it can help you pursue—you might want to explore ways to improve NPS scores, combat reduced customer retention, or understand (and recreate) the success behind a well-received product. The common thread? All these metrics impact your business, and qualitative research can help investigate and improve that impact.
In this article, we’ll take you through seven methods and examples of qualitative research, including when and how to use them.
Conduct qualitative research with Maze, analyze data instantly, and get rich, descriptive insights that drive decision-making.
There are various qualitative UX research methods that can help you get in-depth, descriptive insights. Some are suited to specific phases of the design and development process, while others are more task-oriented.
Here’s our overview of the most common qualitative research methods. Keep reading for their use cases, and detailed examples of how to conduct them.
Method | |||
---|---|---|---|
User interviews | |||
Focus groups | |||
Ethnographic research | |||
Qualitative observation | |||
Case study research | |||
Secondary research | |||
Open-ended surveys | to extract descriptive insights. |
A user interview is a one-on-one conversation between a UX researcher, designer or Product Manager and a target user to understand their thoughts, perspectives, and feelings on a product or service. User interviews are a great way to get non-numerical data on individual experiences with your product, to gain a deeper understanding of user perspectives.
Interviews can be structured, semi-structured, or unstructured . Structured interviews follow a strict interview script and can help you get answers to your planned questions, while semi and unstructured interviews are less rigid in their approach and typically lead to more spontaneous, user-centered insights.
Interviews are ideal when you want to gain an in-depth understanding of your users’ perspectives on your product or service, and why they feel a certain way.
Interviews can be used at any stage in the product design and development process, being particularly helpful during:
💡 A specialized user interview tool makes interviewing easier. With Maze Interview Studies , you can recruit, host, and analyze interviews all on one platform.
Let’s say you’ve designed a recruitment platform, called Tech2Talent , that connects employers with tech talent. Before starting the design process, you want to clearly understand the pain points employers experience with existing recruitment tools'.
You draft a list of ten questions for a semi-structured interview for 15 different one-on-one interviews. As it’s semi-structured, you don’t expect to ask all the questions—the script serves as more of a guide.
One key question in your script is: “Have tech recruitment platforms helped you find the talent you need in the past?”
Most respondents answer with a resounding and passionate ‘no’ with one of them expanding:
“For our company, it’s been pretty hit or miss honestly. They let just about anyone make a profile and call themselves tech talent. It’s so hard sifting through serious candidates. I can’t see any of their achievements until I invest time setting up an interview.”
You begin to notice a pattern in your responses: recruitment tools often lack easily accessible details on talent profiles.
You’ve gained contextual feedback on why other recruitment platforms fail to solve user needs.
A focus group is a research method that involves gathering a small group of people—around five to ten users—to discuss a specific topic, such as their’ experience with your new product feature. Unlike user interviews, focus groups aim to capture the collective opinion of a wider market segment and encourage discussion among the group.
You should use focus groups when you need a deeper understanding of your users’ collective opinions. The dynamic discussion among participants can spark in-depth insights that might not emerge from regular interviews.
Focus groups can be used before, during, and after a product launch. They’re ideal:
The number of participants can make it difficult to take notes or do manual transcriptions. We recommend using a transcription or a specialized UX research tool , such as Maze, that can automatically create ready-to-share reports and highlight key user insights.
You’re a UX researcher at FitMe , a fitness app that creates customized daily workouts for gym-goers. Unlike many other apps, FitMe takes into account the previous day’s workout and aims to create one that allows users to effectively rest different muscles.
However, FitMe has an issue. Users are generating workouts but not completing them. They’re accessing the app, taking the necessary steps to get a workout for the day, but quitting at the last hurdle.
Time to talk to users.
You organize a focus group to get to the root of the drop-off issue. You invite five existing users, all of whom have dropped off at the exact point you’re investigating, and ask them questions to uncover why.
A dialog develops:
Participant 1: “Sometimes I’ll get a workout that I just don’t want to do. Sure, it’s a good workout—but I just don’t want to physically do it. I just do my own thing when that happens.”
Participant 2: “Same here, some of them are so boring. I go to the gym because I love it. It’s an escape.”
Participant 3: “Right?! I get that the app generates the best one for me on that specific day, but I wish I could get a couple of options.”
Participant 4: “I’m the same, there are some exercises I just refuse to do. I’m not coming to the gym to do things I dislike.”
Conducting the focus groups and reviewing the transcripts, you realize that users want options. A workout that works for one gym-goer doesn’t necessarily work for the next.
A possible solution? Adding the option to generate a new workout (that still considers previous workouts)and the ability to blacklist certain exercises, like burpees.
Ethnographic research is a research method that involves observing and interacting with users in a real-life environment. By studying users in their natural habitat, you can understand how your product fits into their daily lives.
Ethnographic research can be active or passive. Active ethnographic research entails engaging with users in their natural environment and then following up with methods like interviews. Passive ethnographic research involves letting the user interact with the product while you note your observations.
Ethnographic research is best suited when you want rich insights into the context and environment in which users interact with your product. Keep in mind that you can conduct ethnographic research throughout the entire product design and development process —from problem discovery to post-launch. However, it’s mostly done early in the process:
While ethnographic studies provide a comprehensive view of what potential users actually do, they are resource-intensive and logistically difficult. A common alternative is diary studies. Like ethnographic research, diary studies examine how users interact with your product in their day-to-day, but the data is self-reported by participants.
⚙️ Recruiting participants proving tough and time-consuming? Maze Panel makes it easy, with 400+ filters to find your ideal participants from a pool of 3 million participants.
You're a UX researcher for a project management platform called ProFlow , and you’re conducting an ethnographic study of the project creation process with key users, including a startup’s COO.
The first thing you notice is that the COO is rushing while navigating the platform. You also take note of the 46 tabs and Zoom calls opened on their monitor. Their attention is divided, and they let out an exasperated sigh as they repeatedly hit “refresh” on your website’s onboarding interface.
You conclude the session with an interview and ask, “How easy or difficult did you find using ProFlow to coordinate a project?”
The COO answers: “Look, the whole reason we turn to project platforms is because we need to be quick on our feet. I’m doing a million things so I need the process to be fast and simple. The actual project management is good, but creating projects and setting up tables is way too complicated.”
You realize that ProFlow ’s project creation process takes way too much time for professionals working in fast-paced, dynamic environments. To solve the issue, propose a quick-create option that enables them to move ahead with the basics instead of requiring in-depth project details.
Qualitative observation is a similar method to ethnographic research, though not as deep. It involves observing your users in a natural or controlled environment and taking notes as they interact with a product. However, be sure not to interrupt them, as this compromises the integrity of the study and turns it into active ethnographic research.
Qualitative observation is best when you want to record how users interact with your product without anyone interfering. Much like ethnographic research, observation is best done during:
You’re conducting UX research for Stackbuilder , an app that connects businesses with tools ideal for their needs and budgets. To determine if your app is easy to use for industry professionals, you decide to conduct an observation study.
Sitting in with the participant, you notice they breeze past the onboarding process, quickly creating an account for their company. Yet, after specifying their company’s budget, they suddenly slow down. They open links to each tool’s individual page, confusingly switching from one tab to another. They let out a sigh as they read through each website.
Conducting your observation study, you realize that users find it difficult to extract information from each tool’s website. Based on your field notes, you suggest including a bullet-point summary of each tool directly on your platform.
Case studies are a UX research method that provides comprehensive and contextual insights into a real-world case over a long period of time. They typically include a range of other qualitative research methods, like interviews, observations, and ethnographic research. A case study allows you to form an in-depth analysis of how people use your product, helping you uncover nuanced differences between your users.
Case studies are best when your product involves complex interactions that need to be tracked over a longer period or through in-depth analysis. You can also use case studies when your product is innovative, and there’s little existing data on how users interact with it.
As for specific phases in the product design and development process:
Your team has recently launched Pulse , a platform that analyzes social media posts to identify rising digital marketing trends. Pulse has been on the market for a year, and you want to better understand how it helps small businesses create successful campaigns.
To conduct your case study, you begin with a series of interviews to understand user expectations, ethnographic research sessions, and focus groups. After sorting responses and observations into common themes you notice a main recurring pattern. Users have trouble interpreting the data from their dashboards, making it difficult to identify which trends to follow.
With your synthesized insights, you create a report with detailed narratives of individual user experiences, common themes and issues, and recommendations for addressing user friction points.
Some of your proposed solutions include creating intuitive graphs and summaries for each trend study. This makes it easier for users to understand trends and implement strategic changes in their campaigns.
Secondary research is a research method that involves collecting and analyzing documents, records, and reviews that provide you with contextual data on your topic. You’re not connecting with participants directly, but rather accessing pre-existing available data. For example, you can pull out insights from your UX research repository to reexamine how they apply to your new UX research objective.
Strictly speaking, it can be both qualitative and quantitative—but today we focus on its qualitative application.
Record keeping is particularly useful when you need supplemental insights to complement, validate, or compare current research findings. It helps you analyze shifting trends amongst your users across a specific period. Some other scenarios where you need record keeping include:
Secondary research is especially valuable when your team faces budget constraints, tight deadlines, or limited resources. Through review mining and collecting older findings, you can uncover useful insights that drive decision-making throughout the product design and development process.
SafeSurf is a cybersecurity platform that offers threat detection, security audits, and real-time reports. After conducting multiple rounds of testing, you need a quick and easy way to identify remaining usability issues. Instead of conducting another resource-intensive method, you opt for social listening and data mining for your secondary research.
Browsing through your company’s X, you identify a recurring theme: many users without a background in tech find SafeSurf ’s reports too technical and difficult to read. Users struggle with understanding what to do if their networks are breached.
After checking your other social media channels and review sites, the issue pops up again.
With your gathered insights, your team settles on introducing a simplified version of reports, including clear summaries, takeaways, and step-by-step protocols for ensuring security.
By conducting secondary research, you’ve uncovered a major usability issue—all without spending large amounts of time and resources to connect with your users.
Open-ended surveys are a type of unmoderated UX research method that involves asking users to answer a list of qualitative research questions designed to uncover their attitudes, expectations, and needs regarding your service or product. Open-ended surveys allow users to give in-depth, nuanced, and contextual responses.
User surveys are an effective qualitative research method for reaching a large number of users. You can use them at any stage of the design and product development process, but they’re particularly useful:
You're a UX researcher for RouteReader , a comprehensive logistics platform that allows users to conduct shipment tracking and route planning. Recently, you’ve launched a new predictive analytics feature that allows users to quickly identify and prepare for supply chain disruptions.
To better understand if users find the new feature helpful, you create an open-ended, in-app survey.
The questions you ask your users:
Most of the responses are positive. Users report using the predictive analytics feature to make last-minute adjustments to their route plans, and some even rely on it regularly. However, a few users find the feature hard to notice, making it difficult to adjust their routes on time.
To ensure users have supply chain insights on time, you integrate the new feature into each interface so users can easily spot important information and adjust their routes accordingly.
💡 Surveys are a lot easier with a quality survey tool. Maze’s Feedback Surveys solution has all you need to ensure your surveys get the insights you need—including AI-powered follow-up and automated reports.
Alongside qualitative research approaches, UX teams also use quantitative research methods. Despite the similar names, the two are very different.
Here are some of the key differences between qualitative research and quantitative research .
Research type | ||||
---|---|---|---|---|
Qualitative research | . | |||
Quantitative research |
Before selecting either qualitative or quantitative methods, first identify what you want to achieve with your UX research project. As a general rule of thumb, think qualitative data collection for in-depth understanding and quantitative studies for measurement and validation.
You’ll often find that knowing the what is pointless without understanding the accompanying why . Qualitative research helps you uncover your why.
So, what about how —how do you identify your 'what' and your 'why'?
The answer is with a user research tool like Maze.
Maze is the leading user research platform that lets you organize, conduct, and analyze both qualitative and quantitative research studies—all from one place. Its wide variety of UX research methods and advanced AI capabilities help you get the insights you need to build the right products and experiences faster.
What is qualitative research?
Qualitative research is a research method that aims to provide contextual, descriptive, and non-numerical insights on a specific issue. Qualitative research methods like interviews, case studies, and ethnographic studies allow you to uncover the reasoning behind your user’s attitudes and opinions.
Can a study be both qualitative and quantitative?
Absolutely! You can use mixed methods in your research design, which combines qualitative and quantitative approaches to gain both descriptive and statistical insights.
For example, user surveys can have both close-ended and open-ended questions, providing comprehensive data like percentages of user views and descriptive reasoning behind their answers.
Is qualitative or quantitative research better?
The choice between qualitative and quantitative research depends upon your research goals and objectives.
Qualitative research methods are better suited when you want to understand the complexities of your user’s problems and uncover the underlying motives beneath their thoughts, feelings, and behaviors. Quantitative research excels in giving you numerical data, helping you gain a statistical view of your user's attitudes, identifying trends, and making predictions.
What are some approaches to qualitative research?
There are many approaches to qualitative studies. An approach is the underlying theory behind a method, and a method is a way of implementing the approach. Here are some approaches to qualitative research:
Run a free plagiarism check in 10 minutes, generate accurate citations for free.
Published on October 26, 2022 by Shona McCombes . Revised on November 21, 2023.
A research question pinpoints exactly what you want to find out in your work. A good research question is essential to guide your research paper , dissertation , or thesis .
All research questions should be:
How to write a research question, what makes a strong research question, using sub-questions to strengthen your main research question, research questions quiz, other interesting articles, frequently asked questions about research questions.
You can follow these steps to develop a strong research question:
The way you frame your question depends on what your research aims to achieve. The table below shows some examples of how you might formulate questions for different purposes.
Research question formulations | |
---|---|
Describing and exploring | |
Explaining and testing | |
Evaluating and acting | is X |
Example research problem | Example research question(s) |
---|---|
Teachers at the school do not have the skills to recognize or properly guide gifted children in the classroom. | What practical techniques can teachers use to better identify and guide gifted children? |
Young people increasingly engage in the “gig economy,” rather than traditional full-time employment. However, it is unclear why they choose to do so. | What are the main factors influencing young people’s decisions to engage in the gig economy? |
Note that while most research questions can be answered with various types of research , the way you frame your question should help determine your choices.
Discover proofreading & editing
Research questions anchor your whole project, so it’s important to spend some time refining them. The criteria below can help you evaluate the strength of your research question.
Criteria | Explanation |
---|---|
Focused on a single topic | Your central research question should work together with your research problem to keep your work focused. If you have multiple questions, they should all clearly tie back to your central aim. |
Answerable using | Your question must be answerable using and/or , or by reading scholarly sources on the to develop your argument. If such data is impossible to access, you likely need to rethink your question. |
Not based on value judgements | Avoid subjective words like , , and . These do not give clear criteria for answering the question. |
Criteria | Explanation |
---|---|
Answerable within practical constraints | Make sure you have enough time and resources to do all research required to answer your question. If it seems you will not be able to gain access to the data you need, consider narrowing down your question to be more specific. |
Uses specific, well-defined concepts | All the terms you use in the research question should have clear meanings. Avoid vague language, jargon, and too-broad ideas. |
Does not demand a conclusive solution, policy, or course of action | Research is about informing, not instructing. Even if your project is focused on a practical problem, it should aim to improve understanding rather than demand a ready-made solution. If ready-made solutions are necessary, consider conducting instead. Action research is a research method that aims to simultaneously investigate an issue as it is solved. In other words, as its name suggests, action research conducts research and takes action at the same time. |
Criteria | Explanation |
---|---|
Cannot be answered with or | Closed-ended, / questions are too simple to work as good research questions—they don’t provide enough for robust investigation and discussion. |
Cannot be answered with easily-found facts | If you can answer the question through a single Google search, book, or article, it is probably not complex enough. A good research question requires original data, synthesis of multiple sources, and original interpretation and argumentation prior to providing an answer. |
Criteria | Explanation |
---|---|
Addresses a relevant problem | Your research question should be developed based on initial reading around your . It should focus on addressing a problem or gap in the existing knowledge in your field or discipline. |
Contributes to a timely social or academic debate | The question should aim to contribute to an existing and current debate in your field or in society at large. It should produce knowledge that future researchers or practitioners can later build on. |
Has not already been answered | You don’t have to ask something that nobody has ever thought of before, but your question should have some aspect of originality. For example, you can focus on a specific location, or explore a new angle. |
Chances are that your main research question likely can’t be answered all at once. That’s why sub-questions are important: they allow you to answer your main question in a step-by-step manner.
Good sub-questions should be:
Here are a few examples of descriptive and framing questions:
Keep in mind that sub-questions are by no means mandatory. They should only be asked if you need the findings to answer your main question. If your main question is simple enough to stand on its own, it’s okay to skip the sub-question part. As a rule of thumb, the more complex your subject, the more sub-questions you’ll need.
Try to limit yourself to 4 or 5 sub-questions, maximum. If you feel you need more than this, it may be indication that your main research question is not sufficiently specific. In this case, it’s is better to revisit your problem statement and try to tighten your main question up.
If you want to know more about the research process , methodology , research bias , or statistics , make sure to check out some of our other articles with explanations and examples.
Methodology
Statistics
Research bias
The way you present your research problem in your introduction varies depending on the nature of your research paper . A research paper that presents a sustained argument will usually encapsulate this argument in a thesis statement .
A research paper designed to present the results of empirical research tends to present a research question that it seeks to answer. It may also include a hypothesis —a prediction that will be confirmed or disproved by your research.
As you cannot possibly read every source related to your topic, it’s important to evaluate sources to assess their relevance. Use preliminary evaluation to determine whether a source is worth examining in more depth.
This involves:
A research hypothesis is your proposed answer to your research question. The research hypothesis usually includes an explanation (“ x affects y because …”).
A statistical hypothesis, on the other hand, is a mathematical statement about a population parameter. Statistical hypotheses always come in pairs: the null and alternative hypotheses . In a well-designed study , the statistical hypotheses correspond logically to the research hypothesis.
Formulating a main research question can be a difficult task. Overall, your question should contribute to solving the problem that you have defined in your problem statement .
However, it should also fulfill criteria in three main areas:
If you want to cite this source, you can copy and paste the citation or click the “Cite this Scribbr article” button to automatically add the citation to our free Citation Generator.
McCombes, S. (2023, November 21). Writing Strong Research Questions | Criteria & Examples. Scribbr. Retrieved August 19, 2024, from https://www.scribbr.com/research-process/research-questions/
Other students also liked, how to define a research problem | ideas & examples, how to write a problem statement | guide & examples, 10 research question examples to guide your research project, "i thought ai proofreading was useless but..".
I've been using Scribbr for years now and I know it's a service that won't disappoint. It does a good job spotting mistakes”
Professor of Anthropology, University of California, Irvine
Roxanne Varzi received funding 20 years ago from fulbright in 2000 and a woodrow wilson in 2002-2004
View all partners
Uncommon Courses is an occasional series from The Conversation U.S. highlighting unconventional approaches to teaching.
Multimodal Anthropology
“Multimodal” simply means using more than academic written text to present research findings.
My first multimodal science project was a third grade project on humpback whales, which was inspired by a vinyl LP recording of their “songs” that National Geographic mailed out to subscribers.
In addition to my written report, I included the whale recording and photographs from National Geographic. I then added a popular piece of material culture by tying the project together with white shoelaces dotted with tiny blue whales.
As an anthropologist today, I use fiction, plays, poetry, sound and film – among other media – to present my findings. To other academics, I’ve often had to defend my methods. But students tell me they love it and want to do the same.
We explore nontraditional ways to present research and discuss why a particular form may or may not work for a particular subject. For example, a student could use the sound of bombs for a research presentation about war, rather than showing violent images, which can demean victims . For a presentation about Iranian underground theater, a fictional story about clandestine rehearsals and relationships , rather than an academic book, could be used to protect subjects’ privacy.
Students learn basic ethnographic research methods, including fly-on-the-wall observation and interviewing. We go out on exploratory research trips around campus to observe, take notes and ask questions. When they choose the topic of their final research projects, they take whatever data they gathered in a given week and present it in different formats, including film, poetry and op-eds.
Practicing multimodal anthropology helps students explore their own strengths and grow their confidence. Not everyone writes well, but some are gifted photographers. Educators have lost the interest of talented students by consistently demanding that research always be presented in term papers.
It also makes the material more accessible to a larger and more diverse audience. Take sound – or even audiobooks – for example. The sound of humpback whales or melting glaciers allows people to experience a place that most cannot actually visit. Audiobooks or spoken-word projects allow readers with vision impairments and reading disabilities to access the research.
There is never just one way to do anything. Students will be successful and can facilitate change in the world by playing to their strengths and by collaborating rather than competing with their peers. They are not their test scores or final grades, but rather a product of the experiences they have with their projects and each other along the way.
I use my own work to explain the process of multimodal anthropology – how and why I did what I did and whether it was successful. I also bring in other guest lecturers, including multimedia artist Kurosh ValaNejad , who put the history of the American-led coup d’etat in Iran in 1953 into a video game .
I am currently developing my own multimodal curriculum in the form of a murder mystery series in which each book explores a different way to present research. My first book, “ Death in a Nutshell: An Anthropology Whodunnit ,” explores visual anthropology – the study of images in culture and the use of images to present culture – and dyslexia.
The course prepares students to practice and learn about multiple research methods. Even if they never write another poem again, studying poetry as an anthropologist sharpens their descriptive abilities. It also leads to useful discussions, like whether poetry, as Plato claims , is too beautiful, alluring and potentially deceptive to be used in what people like to think of as an objective science.
Write an article and join a growing community of more than 188,500 academics and researchers from 5,029 institutions.
Register now
Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.
Scientific Reports volume 14 , Article number: 19140 ( 2024 ) Cite this article
77 Accesses
Metrics details
In recent years, researchers have taken the many-objective optimization algorithm, which can optimize 5, 8, 10, 15, 20 objective functions simultaneously, as a new research topic. However, the current research on many-objective optimization technology also encounters some challenges. For example: Pareto resistance phenomenon, difficult diversity maintenance. Based on the above problems, this paper proposes a many-objective evolutionary algorithm based on three states (MOEA/TS). Firstly, a feature extraction operator is proposed. It can extract the features of the high-quality solution set, and then assist the evolution of the current individual. Secondly, based on Pareto front layer, the concept of “individual importance degree” is proposed. The importance degree of an individual can reflect the importance of the individual in the same Pareto front layer, so as to further distinguish the advantages and disadvantages of different individuals in the same front layer. Then, a repulsion field method is proposed. The diversity of the population in the objective space is maintained by the repulsion field, so that the population can be evenly distributed on the real Pareto front. Finally, a new concurrent algorithm framework is designed. In the algorithm framework, the algorithm is divided into three states, and each state focuses on a specific task. The population can switch freely among these three states according to its own evolution. The MOEA/TS algorithm is compared with 7 advanced many-objective optimization algorithms. The experimental results show that the MOEA/TS algorithm is more competitive in many-objective optimization problems.
Introduction.
In reality, many optimization problems involve multiple conflicting objectives, such as the design of urban public transport routes 1 , production scheduling 2 , securities portfolio management 3 and so on. These types of optimization problems are called multi-objective optimization problems (MOPs). This means that there is no one solution to make all the objectives reach the optimum simultaneously, that is, the optimization of one objective may lead to the deterioration of other objectives 4 , 5 . Consequently, the solutions of MOPs are usually a set of compromise solutions that weigh all objectives. The definition of MOPs is as follows:
Among them, \(f\left(x\right)\) is the m-dimensional objective vector, which contains m conflicting objective functions; \({f}_{i}\left(x\right)\) represents the i-th objective function; x represents the n-dimensional decision variable; \(\Omega \) represents decision space; R m represents the objective space.
In the field of multi-objective optimization, problems with 2 or 3 optimization objectives are called general multi-objective optimization problems (GMOPs). Problems with more than 3 optimization objectives are called many-objective optimization problems (MaOPs) 6 , 7 , 8 . GMOPs aren’t the focus of our attention, as there have been many reports about GMOPs 9 , 10 . On the contrary, MaOPs are the focus of our attention, as there are still some challenges to be solved. The fundamental difference between GMOPs and MaOPs is the number of optimization objectives. Assuming that the number of optimization objectives is m, the probability that one individual dominates another is \(1/{2}^{m-1}\) in theory 11 , 12 . This means that with the increase of the number of optimization objectives, traditional Pareto dominance will fail, Pareto resistance will occur, and most multi-objective optimization algorithms will lose selection pressure in terms of convergence.
In recent years, with the research and exploration of MaOPs, many-objective optimization technology has been developed to a certain extent, and basically 4 mainstream many-objective optimization algorithms have been formed 13 . The first is many-objective optimization algorithm based on dominance. The algorithm modifies the definition of traditional Pareto domination by domination relaxation technique to enhance the selection pressure of the algorithm in terms of convergence. \(\alpha\) -Dominance, \(\upepsilon \) -Dominance and Cone \(\upepsilon \) -Dominance are all common domination relaxation techniques. Compared with traditional Pareto dominance, the effectiveness of dominance relaxation technology has been reported in many works. Therefore, dominance relaxation technology has been widely used to solve MaOPs. However, the current domination relaxation technique also faces two problems: (1) With the increase of the number of optimization objectives, the effect of the domination relaxation technique is getting worse and worse; (2) The domination relaxation technique tends to make the population converge to a certain sub-region of the real Pareto front (PF).
The second is many-objective optimization algorithm based on index. The algorithm guides the selection and evolution of the population by integrating convergence and diversity into one index (such as IGD, HV). Its representative work includes: HypE, MaOEA/IGD, SMS-EMOA. However, the algorithm faces some problems when it is used to solve MaOPs, such as complex index calculation, difficult selection of reference point or reference PF.
The third is many-objective optimization algorithm based on decomposition. The algorithm transforms MaOPs into several single-objective optimization sub-problems through an aggregation function, and then drives the individuals in the neighborhood to update by neighborhood strategy, finally realizes the evolution of the whole population. Its representative work includes: MOEA/D, MOEA/D-D, MOEA/D-DU. However, many-objective optimization algorithm based on decomposition is only suitable for MaOPs with regular PF (such as the DTLZ1 problem). When dealing with MaOPs with irregular PF, many-objective optimization algorithm based on decomposition often performs poorly.
The fourth is many-objective optimization algorithm based on hybrid strategy. The algorithm adopts different search strategies in different environments (different stages or different sub-populations), and uses the advantages of their respective search strategies to deal with complex MaOPs. Its representative work includes: AHM, eMOFEOA, CPSO. In many reports, many-objective optimization algorithm based on hybrid strategy is more suitable for solving MaOPs.
According to the above analysis, this paper considers using the many-objective optimization algorithm based on hybrid strategy, and further proposes the many-objective evolutionary algorithm based on three states (MOEA/TS). The innovations and contributions of this paper are as follows: (1) A feature extraction operator is proposed. The feature extraction operator is a feature extractor, which can extract the features of the high-quality solution set, and then assist the evolution of the current individual. (2) Based on the Pareto front layer, the concept of “individual importance degree” is proposed. The importance degree of an individual can reflect the importance of the individual in the same Pareto front layer, so as to further distinguish the advantages and disadvantages of different individuals in the same front layer, and effectively solve the phenomenon of Pareto resistance. (3) A repulsion field method is proposed. The repulsion field is used to maintain the diversity of the population in the objective space, so that the population can be evenly distributed on the real PF. (4) Design a new concurrent algorithm framework. In the framework, the algorithm is divided into three states, and each state focuses on a specific task. The population can freely switch among these three states according to its own evolution.
The remainder of this paper is organized as follows: Sect. " Preparatory work " introduces the basic definition, related work and research motivation. Sect. " Basic definition " introduces each part of the MOEA/TS algorithm in detail. Sect. " Related work " introduces the test results of MOEA/TS algorithm and 7 advanced many-objective optimization algorithms on various test problems, and then analyzes and summarizes them according to the test results. Sect. " Many-objective optimization algorithm based on dominance "summarizes this article and looks forward to future work.
Basic definition.
In this section, we will introduce some basic definitions related to many-objective optimization technology.
Definition of dominance: if solution x isn’t worse than solution y in all objectives and solution x is better than solution y in at least one objective, it is said that x dominates y . That is, if \(\forall i\in \left\{\text{1,2},3,...,m\right\}\) satisfies \({f}_{i}\left(x\right)\le {f}_{i}\left(y\right)\) and \(\exists j\in \left\{\text{1,2},3,...,m\right\}\) satisfies \({f}_{j}\left(x\right)<{f}_{j}\left(y\right)\) , it is said that x dominates y .
Definition of non-dominated solution: if there are no solutions that can dominate x in the decision space, then x is called a Pareto optimal solution or a non-dominated solution. That is, if \(\nexists {x}^{*}\in \Omega \) makes x* dominate x , then x is called a Pareto optimal solution or a non-dominated solution.
Definition of Pareto optimal solution set: the set composed of Pareto optimal solutions is called the Pareto optimal solution set (PS). The mathematical description of PS is as follows:
Definition of Pareto front: the mapping of PS in the objective space is called Pareto front (PF). The mathematical description of PF is as follows:
The goal of the many-objective optimization technology is to find a set of non-dominated solutions that are close to the real PF (convergence) and make them well distributed on the real PF (diversity).
In recent years, many scholars have conducted in-depth research and exploration in the many-objective optimization technology.
Considering the limitations of Pareto dominance relationship in high-dimensional objective space, Zhou et al 14 proposed a many-objective optimization algorithm based on dominance relation selection. Firstly, they introduced an angle domination relationship with higher selection pressure based on the traditional Pareto domination relationship, and designed a new dominance selection strategy. Additionally, they proposed an angle-based individual distribution method to ensure even population distribution in the objective space. The algorithm shows strong competitiveness in solving MaOPs. Wang et al 15 believed that as the number of objectives increased, the traditional dominance relationship would become invalid. Therefore, they proposed a modified dominance relation. That is, they used penalty-based adaptive matrix regions to assist the traditional dominance relationship. Further, for MaOPs with irregular Pareto fronts, they introduced a population-based adaptive adjustment method to replace the predefined weight vector. On this basis, for MaOPs, they developed a many-objective optimization algorithm based on modified dominance relation and adaptive adjustment method. Zhang et al 16 believed that the current many-objective optimization algorithms focused too much on convergence, which would cause the population to converge to a certain sub-region of the real Pareto front. In order to solve this problem, they proposed a many-objective optimization algorithm based on double distance domination. In this algorithm, double distance can not only measure the convergence of the algorithm to adapt to different Pareto fronts, but also combine angle-based niche technology to emphasize the diversity of the algorithm. In addition, they also designed a special mutation operator. This operator can generate high-quality individuals in sparse areas to improve the diversity of the algorithm.
Aiming at the high complexity problem of hypervolume computation, Shang et al 17 proposed a new multi-objective evolutionary algorithm (MOEA) based on R2 index, namely the R2HCA-EMOA algorithm. The core idea of this algorithm is to use R2 index variables to approximate the contribution of hypervolume. The basic framework of the proposed algorithm is similar to that of SMS-EMOA. In order to improve the calculation efficiency of the algorithm, the utility tensor structure is introduced to calculate R2 index variables. In addition, the normalization mechanism is incorporated into the R2HCA-EMOA algorithm to improve its performance. Zhang et al 18 believed that the loss of selection pressure was the core reason for the poor performance of the algorithm. In order to solve this problem, they proposed a many-objective optimization algorithm based on fitness evaluation and hierarchical grouping. The fitness evaluation method combined the convergence measure based on the cos function and the diversity measure based on angle to create the selection pressure of convergence and diversity. In order to further strengthen the selection pressure, they proposed a hierarchical grouping strategy. Firstly, individuals are divided into different layers by front index, and then individuals in the same layer are divided into different groups by R2 index. Although some indexes can approximate the contribution of HV, However, Nan et al 19 believed that the key of performance evaluation was to find the worst solution rather than accurately approaching the HV value of each solution. In order to improve the ability to identify the worst solution, they proposed a two-stage R2 index evaluation method. In the first stage, the R2 indexes of all individuals are roughly evaluated to select some candidate solutions. In the second stage, these candidate solutions are accurately evaluated. Finally, they proposed a many-objective optimization algorithm based on the two-stage R2 index.
In order to balance the convergence and diversity of the decomposition-based algorithm and reduce its dependence on the real PF direction, Wu et al 20 developed a many-objective optimization algorithm based on antagonistic decomposition method. This method utilizes the complementary characteristics of different sub-problems in a single example. Specifically, two populations are co-evolved by two sub-problems with different contours and opposite search directions. In order to avoid allocating redundant computing resources to the same area of PF, two populations are matched into one-to-one pairing according to their working areas on PF. In mating selection, each solution pair can only contribute one parent at most. In order to improve the performance of decomposition-based algorithms, Fan et al 21 proposed a differential multi-objective optimization algorithm based on decomposition. Firstly, they designed a neighborhood intimacy factor to improve the diversity of the algorithm based on the characteristics of neighborhood search. Then, they introduced a Gaussian mutation operator with dynamic step size to enhance the algorithm’s ability to escape from local optimal regions and improve convergence. Finally, they combined a difference strategy with the decomposition-based multi-objective optimization algorithm to further strengthen its evolutionary ability. Peng et al 22 believed that data dimensionality reduction could be applied to the objective space. Based on this consideration, they proposed a many-objective optimization algorithm based on projection. Firstly, they used the idea of data dimensionality reduction and spatial decomposition to divide the objective space into projection plane and free dimension. Then, a double elite strategy was used to maintain the balance between convergence and diversity of the algorithm. Finally, the algorithm based on decomposition was used as the algorithm of free dimension to solve MaOPs.
Aiming at convergence problem and diversity problem of the algorithm, Sun et al 23 proposed a many-objective optimization algorithm based on two independent stages. The algorithm deals with convergence and diversity problems in two independent and successive stages. Firstly, they introduced a non-dominated dynamic weight aggregation method, which is capable of identifying the Pareto optimal solutions of MaOPs. Then, they used these solutions to learn the Pareto optimal subspace in order to solve the convergence problem. Finally, the diversity problem was solved by using reference lines in the Pareto optimal subspace. Considering the advantages of the multi-objective and multi-population (MPMO) framework in solving MaOPs, Yang et al 24 proposed an algorithm based on the MPMO framework. The algorithm adopts the deviation sorting (BS) method to solve MaOPs, so as to obtain good convergence and diversity. In terms of convergence, the BS method is applied to each population in the MPMO framework, and the effect of non-dominant sorting is enhanced by the optimization objectives of the corresponding population. In terms of diversity, the maintenance method based on reference vector is used to save the diversity solutions. Aiming at the five-objective job shop scheduling problem (JSSP), Liu et al 25 proposed a new genetic algorithm based on the MPMO framework. Firstly, five populations are used to optimize five objectives, respectively. Secondly, in order to prevent each population from focusing only on its corresponding single objective, an archive sharing technology (AST) is proposed to store the elite solutions collected from five populations, so that the population can obtain the optimization information of other objectives from the archive. Thirdly, the archive updating strategy (AUS) is proposed to further improve the quality of the solutions in the archive.
Based on the related work, we believe that there are still the following problems in the current many-objective optimization technology:
(1) The diversity and convergence of the algorithm are difficult to balance. Most algorithms can’t coordinate the balance between them well, and they either emphasize convergence or diversity too much, which leads to poor quality of the non-dominated solution set.
(2) It is difficult to maintain the convergence of the algorithm. When the number of optimization objectives is large, the algorithm will produce Pareto resistance, and the traditional Pareto dominance may fail.
(3) It is difficult to maintain the diversity of the algorithm. Especially when the real PF is complex or the latitude of the objective space is high, individuals may have the clustering effect, and the population may not be evenly distributed on the real PF.
(4) The evolution efficiency of the algorithm is low. The traditional evolution operators have strong randomness and low evolution efficiency, and aren’t suitable for dealing with MaOPs.
Therefore, solving these problems and providing a good many-objective optimization algorithm constitute the research motivation of this paper.
For problem 1, some work attempts to separate the convergence optimization and diversity optimization of the algorithm, thus designing a concurrent algorithm architecture. Concurrent algorithm architecture means that only one of convergence or diversity is considered in one iteration instead of considering both convergence and diversity simultaneously. In order to solve GMOPs, Professor Ye Tian 26 tried to design a concurrent algorithm architecture and proposed the MSEA algorithm, and the experimental results were satisfactory. Therefore, it seems to be a feasible path to solve MaOPs by using concurrent algorithm architecture. However, recent research 23 shows that in MaOPs, the concurrent algorithm architecture seems to be unstable, and the experimental results fluctuate greatly (such as MaOEA/IT algorithm). Because when the algorithm only considers the convergence of the population, it often affects the diversity of the population; Similarly, when the algorithm only considers the diversity of the population, it often affects the convergence of the population. If a coordination intermediary can be added to the concurrent algorithm architecture to alleviate the contradiction between diversity and convergence, the concurrent algorithm architecture will become stable and its superiority will be truly reflected. Based on this motivation, this paper proposes a new concurrent algorithm framework. In the new algorithm framework, the algorithm is divided into three states, namely, convergence maintenance state, diversity maintenance state and coordination state. Each state focuses on a specific task. That is, the convergence maintenance state is responsible for improving the population convergence; Diversity maintenance state is responsible for improving population diversity; the coordination state is responsible for coordinating the contradiction between diversity and convergence. The population can freely switch among these three states according to its own evolution.
For problem 2, some scholars try to modify the definition of traditional Pareto dominance by using dominance relaxation technology to enhance the selection pressure of the algorithm in terms of convergence. However, with the increase of the number of optimization objectives, the effect of dominance relaxation technology is getting worse and worse. They only focus on the modification of Pareto domination definition, but ignore the difference between objective values. If we can distinguish the importance of different individuals by using the difference between the objective values, we can further create the selection pressure of the algorithm in terms of convergence, and finally Pareto resistance will be eliminated. Therefore, based on Pareto front layer, this paper proposes the concept of “individual importance degree”. The importance degree of an individual can reflect the importance of the individual in the same Pareto front layer, so as to further distinguish the advantages and disadvantages of different individuals in the same front layer, and effectively solve the phenomenon of Pareto resistance. Obviously, compared with domination relaxation technique, individual importance degree has greater advantages.
For problem 3, the traditional diversity maintenance technology isn’t suitable for high-dimensional objective space. For instance: the niche method, the density evaluation method, and the weight vector method. In the field of microphysics, when the distance between particles is too close, repulsion will push the particles away from their neighbors. On the contrary, when the distance between particles is too great, the repulsion will decrease and the particles tend to be close to the neighboring particles. This way makes the distribution of particles present a state of mutual coordination. Based on the characteristics of particle distribution, a repulsion field method is proposed in this paper. The repulsion field is used to maintain the diversity of the population in the objective space, so that the population can be evenly distributed on the real PF.
For problem 4, traditional evolution operators aren’t suitable for dealing with MaOPs. Because traditional evolution operators have strong randomness and low evolution efficiency. For instance: the binary crossover operator, the polynomial mutation operator, and the differential evolution operator. In principal component analysis, the decomposition of the covariance matrix and correlation matrix is a very important step. By decomposing the covariance matrix or the correlation matrix, we can obtain a set of orthogonal bases. These orthogonal bases are the most important features of the original data 27 . Therefore, this paper designs a feature extraction operator based on Cholesky decomposition 28 . The feature extraction operator can be understood as a feature extractor. It can extract the features of the high-quality solution set, and then assist the evolution of the current individual. Obviously, compared with traditional evolution operators, the feature extraction operator has higher evolution efficiency.
Feature extraction operator.
The feature extraction operator is a feature extractor, which can extract the features of the high-quality solution set, and then assist the evolution of the current individual. The workflow of the feature extraction operator is shown in Fig. 1 .
The workflow of feature extraction operator.
In the first step, W high-quality solutions \({x}^{1},{x}^{2},{x}^{3},...,{x}^{W}\) are selected from the population. These W solutions will form the high-quality solution set S.
In the second step, calculate the mean \(\overline{x}\) and covariance matrix A of the high-quality solution set S:
Among them, \({x}^{i}={\left({x}_{1}^{i},{x}_{2}^{i},{x}_{3}^{i},...,{x}_{n}^{i}\right)}^{T}, i\in (1,...,W); {x}_{j}={\left({x}_{j}^{1},{x}_{j}^{2},{x}_{j}^{3},...,{x}_{j}^{W}\right)}^{T}, j\in (1,...,n)\)
In the third step, Cholesky decomposition is performed on the covariance matrix A. That is, the covariance matrix A is decomposed into the product of the lower triangular matrix and the transposition of the lower triangular matrix. Assuming that the lower triangular matrix is L, there is
Through formula \(A=L*{L}^{T}\) , we can calculate \({a}_{11}={l}_{11}^{2}\) , that is, \({l}_{11}=\sqrt{{a}_{11}}\) . Then, according to \({a}_{i1}={l}_{i1}*{l}_{11}\) , we can get \({l}_{i1}={a}_{i1}/{l}_{11}\) , so we can get the first column element of matrix L .
Assuming that we have calculated the first k-1 column elements of the matrix L. Through
In this way, we can solve the k-th column element of matrix L through the first k-1 column elements of matrix L. Then, we can solve matrix L by recursion.
In the fourth step, the sampling vector \(s={\left({s}_{1},...,{s}_{n}\right)}^{T}\) is generated by Gaussian distribution \(N\left(\text{0,0.7}\right)\) . Then, a feature solution is generated.
Among them, \({x}^{feature}={\left({x}_{1}^{feature},...,{x}_{n}^{feature}\right)}^{T}\)
It should be noted that the standard deviation std is an important parameter of the Gaussian distribution. In this paper, the standard deviation std is set to 0.7. The parameter analysis verifies that 0.7 is a reasonable standard deviation. For more details on parameter analysis, please browse the experiment chapter (Parameter sensitivity analysis section).
In the fifth step, assuming that the selected individual is \({x}^{i}({x}_{1}^{i},...,{x}_{n}^{i})\) . Based on binary crossover operator 29 and feature solution, the formula of generating offspring individual is as follows:
Among them, \({c({c}_{1},...,{c}_{n})}^{T}\) is the offspring individual. \({\beta }_{k}\) is dynamically determined by the feature factor \(\mu \) :
Among them, \(rand\) is used to generate a random number between 0 and 1; \(r and i(\text{0,1})\) is used to generate 0 or 1 randomly.
For the design principle of formula ( 15 ), please browse the Supplementary Information Document.
In the sixth step, the individual \(c\) is detected and repaired. When some components in individual \(c\) exceed the upper bound or lower bound, these components need to be repaired. The repair formula is as follows:
Among them, \({c}_{i}^{u}\) , \({c}_{i}^{l}\) represent the upper bound and lower bound of the i-th component of individual \(c\) , respectively. \({{c}{\prime}({c}_{1}{\prime},...,{c}_{n}{\prime})}^{T}\) represents the repaired individual.
When the number of optimization objectives is large, the algorithm will produce Pareto resistance, and the traditional Pareto dominance may fail. Some scholars try to modify the definition of traditional Pareto dominance by using dominance relaxation technology to enhance the selection pressure of the algorithm in terms of convergence. However, with the increase of the number of optimization objectives, the effect of dominance relaxation technology is getting worse and worse. They only focus on the modification of Pareto domination definition, but ignore the difference between objective values. Figure 2 shows 4 non-dominant individuals. Among them, individual B is the closest to Origin O, individual C is second, individual A is third, and individual D is the farthest from Origin O. This means that in the population, individual B is the most important, individual C is the second most important, individual A is the third most important, and individual D is the least important. In addition, we can also find from Fig. 2 that there is a significant difference between the objective values of individual B and the objective values of other individuals, that is, \(\sum_{X\in \left\{A,C,D\right\}}\sum_{i=1}^{2}{f}_{i}(B)-{f}_{i}(X)\) is the smallest. This shows that there is a special relationship between the importance of individuals and the difference of the objective values. Based on this discovery, if we can distinguish the importance of different individuals by using the difference between the objective values, we can further create the selection pressure of the algorithm in terms of convergence, and finally Pareto resistance will be eliminated. Therefore, based on Pareto front layer, we propose the concept of “individual importance degree”.
Schematic diagram of individual importance.
Assuming that there are n solutions in a certain Pareto front layer, the objective function values of these solutions are normalized to [0,1] based on the maximum and minimum values of each objective function. \({f}_{k}{\prime}({x}^{i})\) represents the k-th normalized objective function value of individual \({x}^{i}\) .
Define the Pareto dominance function
The trend of the Pareto dominance function is shown in Fig. 3 .
The trend of the Pareto dominance function.
Pareto dominance function can be used to reflect the dominance degree among different individuals. For example, \(PDF({f}_{k}{\prime}({x}^{i})-{f}_{k}{\prime}({x}^{j}))\) represents the dominance degree of individual \({x}^{i}\) to individual \({x}^{j}\) on the k-th objective function; \(PDF({f}_{k}{\prime}({x}^{j})-{f}_{k}{\prime}({x}^{i}))\) represents the dominance degree of individual \({x}^{j}\) to individual \({x}^{i}\) on the k-th objective function; Obviously, the greater the dominance degree, the better one individual is than another on one objective function. Therefore, on one objective function, the dominance degree of one individual to another can be expressed as:
On this basis, the dominance degree of one individual to another can be expressed as:
Further, the importance degree of one individual to another can be expressed as:
Importance degree can indicate the importance of one individual to another. The greater the importance degree, the more important one individual is than another.
Since a certain Pareto front layer has n solutions, each solution needs to be compared with other n-1 solutions, so as to construct n-1 competing pairs. Assuming that an individual is \({x}^{i}\) , then the n-1 competing pairs are \(\left({x}^{i},{x}^{1}\right),\left({x}^{i},{x}^{2}\right),...,\left({x}^{i},{x}^{j}\right),...,\left({x}^{i},{x}^{n}\right)\) , respectively (note: \(i\ne j\) ). Thus, the importance degree of individual \({x}^{i}\) to the other n-1 individuals is \(Imp\left({x}^{i},{x}^{1}\right),Imp\left({x}^{i},{x}^{2}\right),...,Imp\left({x}^{i},{x}^{j}\right),...,Imp\left({x}^{i},{x}^{n}\right)\) , respectively (note: \(i\ne j\) ).
Finally, the importance degree of the individual \({x}^{i}\) can be expressed as:
The importance degree of one individual can reflect the importance of the individual in the same Pareto front layer, so as to further distinguish the advantages and disadvantages of different individuals in the same front layer. The greater the importance degree of one individual, the more important it is in the same Pareto front layer.
Figure 4 shows the use of individual importance degree. Firstly, based on a certain Pareto front layer, the competition pools and competition pairs are constructed. Then, the individual importance degree of different individuals is calculated by Formula ( 24 ). Finally, the importance of different individuals in the same Pareto front layer is obtained.
The use of individual importance degree.
Taking the two-objective optimization problem as an example, it is assumed that there are 4 non-dominant individuals. They are \(A\left(\text{17,5}\right)\) , \(B\left(\text{9,7}\right)\) , \(C\left(\text{7,15}\right)\) and \(D\left(\text{5,25}\right)\) , respectively. It means that these 4 individuals belong to the first non-dominant layer, and their advantages and disadvantages can’t be compared by the non-dominant rank. The distribution of 4 individuals in the objective space is shown in Fig. 5 (a).
The distribution of 4 individuals.
In order to better compare the advantages and disadvantages of these 4 individuals, we use the individual importance degree to deal with these 4 individuals. Firstly, the objective function values of these 4 individuals are normalized to [0,1]. After normalization, the coordinates of these 4 individuals are \(A\left(\text{1,0}\right)\) , \(B\left(\text{0.333,0.1}\right)\) , \(C\left(\text{0.167,0.5}\right)\) and \(D\left(\text{0,1}\right)\) , respectively. The distribution of 4 individuals in the normalized objective space is shown in Fig. 5 (b). Next, according to Fig. 4 , the competition pools and competition pairs are constructed. Then, according to formula ( 22 ) and formula ( 23 ), the \(P\left({x}^{i},{x}^{j}\right)\) and \(Imp\left({x}^{i},{x}^{j}\right)\) of each competition pair are calculated. The calculation results are shown in Table 1 . Finally, according to the formula ( 24 ), the importance degree of these 4 individuals is 0.1918, 0.9488, 0.6673 and 0.1921, respectively. The results show that individual B is the most important, individual C is the second most important, individual D is the third most important, and individual A is the least important. This result is consistent with the intuitive perception that we get from Fig. 5 (b). Based on the above example, we believe that the concept of individual importance degree and related process are effective and can achieve the desired goals.
In the field of microphysics, when the distance between particles is too close, repulsion will push the particles away from their neighbors. On the contrary, when the distance between particles is too great, the repulsion will decrease and the particles tend to be close to the neighboring particles. This way makes the distribution of particles present a state of mutual coordination (As shown in Fig. 6 ). Based on the characteristics of particle distribution, a repulsion field method is proposed in this paper. The repulsion field is used to maintain the diversity of the population in the objective space, so that the population can be evenly distributed on the real PF.
The uniform distribution of microscopic particles.
Firstly, according to the maximum and minimum values of each objective function, the objective function values of all solutions in the population are normalized to [0,1]. \({f}_{k}{\prime}({x}^{i})\) represents the k-th normalized objective function value of individual \({x}^{i}\) .
Then, a repulsion potential field with repulsion radius r is constructed around each individual. Assuming that a repulsion potential field has been constructed for individual \({x}^{i}\) , then all individuals within the repulsion potential field will be subject to the repulsion potential from individual \({x}^{i}\) . The magnitude of the repulsion potential depends on the distance between other individuals and individual \({x}^{i}\) . When other individuals are outside the repulsion potential field of individual \({x}^{i}\) , the repulsion potential is 0. When other individuals are within the repulsion potential field of individual \({x}^{i}\) , the closer the other individuals are to individual \({x}^{i}\) , the greater the repulsion potential that they obtain. Assuming that there is individual \({x}^{j}\) , then the repulsion potential that individual \({x}^{j}\) obtains is
Among them, \(\rho \) is the gain coefficient of the repulsion potential field, usually set to 1; \(r\) is the radius of the repulsion potential field; \(dis\left({x}^{j},{x}^{i}\right)\) represents the euclidean distance between individual \({x}^{j}\) and individual \({x}^{i}\) in the objective space. The formula is as follows:
Further, the repulsion \(Rep\left({x}^{j},{x}^{i}\right)\) that individual \({x}^{j}\) obtains is the negative gradient of the repulsion potential \(Repfield\left({x}^{j},{x}^{i}\right)\) . The formula is as follows:
It means that when \(dis\left({x}^{j},{x}^{i}\right)\le r\) , the smaller \(dis\left({x}^{j},{x}^{i}\right)\) is, the larger \(Rep\left({x}^{j},{x}^{i}\right)\) is. when \(dis\left({x}^{j},{x}^{i}\right)>r\) , \(Rep\left({x}^{j},{x}^{i}\right)=0\) .
Based on the repulsion potential field, the total repulsion potential that individual \({x}^{j}\) obtains is
Finally, the total repulsion that individual \({x}^{j}\) obtains is
It should be noted that the repulsion potential and repulsion proposed in this paper are both vectors. It means that repulsion potential and repulsion have both magnitude and direction. The addition of different repulsion is the vector synthesis of repulsion, rather than the pure numerical addition. This is also an obvious feature that the repulsion field method is different from other scalar function methods (such as niche method). Figure 7 shows the vector synthesis process of repulsion in a two-dimensional space environment. Among them, F SUM is the total repulsion that individual A obtains; F BA is the repulsion generated by individual B to individual A; F CA is the repulsion generated by individual C to individual A.
The vector synthesis process of repulsion.
In the repulsion field method, the individual with large repulsion usually means that the individual is located in the multiple repulsion potential field that other individuals construct. It indicates that the individual is located in a dense area in the objective space and is close to other individuals. Therefore, individuals with large repulsion aren’t conducive to maintaining population diversity. Naturally, we hope that individuals with large repulsion can move away from dense areas in the objective space along the direction of repulsion. Based on this idea, firstly, we need to find some individuals closest to the direction of the repulsion to construct a high-quality solution set. Then, the feature extraction operator is used to extract the location features of the high-quality solution set. Finally, based on these features, individuals with large repulsion can evolve along the direction of repulsion. As shown in Fig. 8 , individual D and individual E are the individuals closest to the direction of repulsion. The feature extraction operator is used to extract the position features of these two individuals. Based on these features, individual A evolves into individual A*, which is far away from the previous dense area.
Individual A is far away from the dense area.
It should be noted that the feature extraction operator has the randomness caused by Gaussian sampling. Therefore, the evolution of individuals also has a certain degree of randomness.
The framework of the MOEA/TS algorithm is shown in Fig. 9 . Firstly, the relevant parameters of the algorithm are initialized; secondly, judge which state the algorithm is in. If the algorithm is in the convergence maintenance state, the following steps are adopted to improve the convergence of the algorithm: (1) Randomly select the parent individual. (2) Use feature extraction operator to generate offspring individuals. (3) If the offspring individual is superior to the individual with the worst convergence in the population, the worst individual is replaced by the offspring individual. If the algorithm is in the diversity maintenance state, the following steps are adopted to improve the diversity of the algorithm: (1) Select the individual with the worst diversity in the population. (2) Use feature extraction operator to generate offspring individuals. (3) If the offspring individual is superior to the individual with the worst diversity in the population, the worst individual is replaced by the offspring individual. If the algorithm is in the coordination state, the following steps are adopted to coordinate the convergence and diversity of the algorithm: (1) Randomly select the parent individual. (2) Use the Gaussian mutation operator to generate offspring individuals. (3) If the offspring individual is superior to the parent individual in convergence and diversity, the parent individual is replaced by the offspring individual. Then, it is judged whether the algorithm has completed the i-th iteration. If the algorithm doesn’t complete the i-th iteration, the corresponding maintenance step or coordination step is re-executed. If the algorithm completes the i-th iteration, the current state of the algorithm is updated. Finally, it is judged whether the algorithm ends. If the algorithm doesn’t end, the corresponding maintenance step or coordination step is performed according to the current state of the algorithm. If the algorithm is finished, the population is output.
The framework of MOEA/TS algorithm.
Main framework.
This section describes the main framework of the MOEA/TS algorithm. The pseudo-code of the main framework of the MOEA/TS algorithm is shown in Algorithm 1. The main steps include: in line (1), initializing population P, repulsion field radius r, and state value (state=1 means that the algorithm is in convergence maintenance state, state=2 means that the algorithm is in diversity maintenance state, and state=3 means that the algorithm is in coordination state.); In line (2), the Front value, Imp value and Rep value of each solution in population P are calculated (The Front value is calculated by the fast non-dominated sorting method.); In line (3), it is judged whether the algorithm meets the termination condition (The termination condition is usually the maximum iterations.); In line (4), the count value is initialized. The count value is used to count the number of updated solutions in the i-th iteration; in lines (5)-(11), according to the current state of the algorithm, the update way of the population is selected. When state=1, the convergence of the population is updated. When state=2, the diversity of the population is updated. When state=3, the convergence and diversity of the population are coordinated; in line (12), the state value of the algorithm is updated according to the current state of the algorithm and the count value.
Main framework.
This section mainly describes the convergence maintenance of the population. The pseudo-code of convergence maintenance is shown in Algorithm 2. The main steps include: in line (1), the algorithm enters the i-th iteration; In line (2), one parent individual is randomly selected from population P; In lines (3)–(4), based on Front, Imp, the high-quality solution set S is constructed; In line (5), the feature extraction operator is used to extract the features of the high-quality solution set S, and then assist the evolution of the parent individual; In line (6), the individual with the worst convergence in the population is found; In lines (7)–(13), if the offspring individual is superior to the individual with the worst convergence in the population, the worst individual is replaced by the offspring individual and flag is marked as 1. If the offspring individual is inferior to the individual with the worst convergence in the population, the flag is marked as 0. Among them, the flag value is used to indicate whether the population P has changed (flag=0 means that the population P hasn’t changed, flag=1 means that the population P has changed.); In lines (14)–(16), it is judged whether flag equals 1. If flag equals 1, the count value is updated, and the Front value, Imp value and Rep value of each solution in population P are updated.
Convergence maintenance.
This section mainly describes the diversity maintenance of the population. The pseudo-code of diversity maintenance is shown in Algorithm 3. The main steps include: in line (1), the algorithm enters the i-th iteration; In line (2), the individual with the worst diversity in the population is found; In line (3), according to the direction of total repulsion that the worst individual obtains, the distance \({dis}_{j}\) can be calculated; In line (4), based on \({dis}_{j}\) , the high-quality solution set S is constructed; In line (5), the feature extraction operator is used to extract the features of the high-quality solution set S, and then assist the evolution of the worst individual; In lines (6)–(12), if the offspring individual is superior to the individual with the worst diversity in the population, the worst individual is replaced by the offspring individual and flag is marked as 1. If the offspring individual is inferior to the individual with the worst diversity in the population, the flag is marked as 0; In lines (13)–(15), it is judged whether flag equals 1. If flag equals 1, the count value is updated, and the Front value, Imp value and Rep value of each solution in population P are updated.
Diversity maintenance.
This section mainly describes the coordination of convergence and diversity of the population. The pseudo-code of coordination of convergence and diversity is shown in Algorithm 4. The main steps include: in line (1), the algorithm enters the i-th iteration; In line (2), one parent individual is randomly selected from population P; In line (3), based on the parent individual, the Gaussian mutation operator is used to generate the offspring solution; In lines (4)–(10), if the offspring individual is superior to the parent individual in convergence and diversity, the parent individual is replaced by the offspring individual and flag is marked as 1. If the offspring individual is inferior to the parent individual in convergence and diversity, the flag is marked as 0; In lines (11)–(13), it is judged whether flag equals 1. If flag equals 1, the count value is updated, and the Front value, Imp value and Rep value of each solution in population P are updated.
Coordination.
This section mainly describes the feature extraction operator. The pseudo-code of the feature extraction operator is shown in Algorithm 5. The main steps include: in line (1), the features \(\overline{x},L\) of the high-quality solution set S are extracted by formula ( 4 ) and formula ( 13 ); In line (2), the sampling vector \(s={\left({s}_{1},...,{s}_{n}\right)}^{T}\) is generated by the Gaussian distribution \(N\left(\text{0,0.7}\right)\) ; In line (3), based on \(\text{s},\overline{x},L\) , the feature solution \({x}^{feature}\) is generated by formula ( 14 ); In line (4), based on parent, \({x}^{feature}\) , the offspring solution O is generated by formula ( 15 ).
Feature extraction.
In this paper, the algorithm state is further updated according to the current state of the algorithm and the stability of the population. The pseudo-code of the update of the algorithm state is shown in Algorithm 6. When the algorithm is in the convergence maintenance state and the number of updated solutions in the i-th iteration is less than or equal to 5%*N, it is considered that the population tends to be stable in terms of convergence, then the algorithm turns to the diversity maintenance state; When the algorithm is in the diversity maintenance state and the number of updated solutions in the i-th iteration is less than or equal to 5%*N, it is considered that the population tends to be stable in terms of diversity, then the algorithm turns to the coordination state; When the algorithm is in the coordination state and the number of updated solutions in the i-th iteration is less than or equal to 5%*N, it is considered that the population tends to be stable in terms of coordination, then the algorithm turns to the convergence maintenance state. It should be noted that the threshold value T is a key parameter in measuring whether the population tends to be stable or not. In this paper, the threshold value T is set to 5%. The parameter analysis verifies that 5% is a reasonable threshold. For more details on parameter analysis, please browse the experiment chapter (Parameter sensitivity analysis section).
Determination state.
Assuming that the size of the population is N, the number of the objective function is m, the dimension of the decision variable is n, and the size of the high-quality solution set is W, then the computational complexity of Rep is O(mN2), the computational complexity of Front is O(mN2), and the computational complexity of Imp is O(mN2). The core steps of the feature extraction operator (Algorithm 5) include the construction of the covariance matrix and Cholesky decomposition. The computational complexity of covariance matrix construction is O(Wn2) and the computational complexity of Cholesky decomposition is O(n3/6). Therefore, the computational complexity of the feature extraction operator (Algorithm 5) is O(Wn2+n3/6). The core steps of convergence maintenance (Algorithm 2) include population ranking, feature extraction operator, selection of the worst individual, and updating of Front, Imp and Rep. Their computational complexity is O(N2), O(Wn2+n3/6), O(N), O(mN2), O(mN2), O(mN2), respectively. Therefore, the computational complexity of convergence maintenance (Algorithm 2) is O(N(N2+Wn2+n3/6+N+3mN2)). The core steps of diversity maintenance (Algorithm 3) include selection of the worst individual, distance calculation, population ranking, feature extraction operator, and updating of Front, Imp and Rep. Their computational complexity is O(N), O(nN), O(N2), O(Wn2+n3/6), O(mN2), O(mN2), O(mN2), respectively. Therefore, the computational complexity of diversity maintenance (Algorithm 3) is O(N(N+nN+N2+Wn2+n3/6+3mN2)). The core steps of coordination of convergence and diversity (Algorithm 4) include the Gaussian mutation operator, and updating of Front, Imp and Rep. Their computational complexity is O(n), O(mN2), O(mN2), O(mN2), respectively. Therefore, the computational complexity of coordination of convergence and diversity (Algorithm 4) is O(N(n+3mN2)). The computational complexity of Determination State (Algorithm 6) is O(1). Based on the above computational complexity analysis, the computational complexity of one iteration of the MOEA/TS algorithm is max{O(N(N2+Wn2+n3/6+N+3mN2)), O(N(N+nN+N2+Wn2+n3/6+3mN2)), O(N(n+3mN2))}+O(1)≈max{O(NWn2+Nn3+mN3), O(NWn2+Nn3+mN3), O(mN3)}= O(NWn2+Nn3+mN3). In this paper, N>>max{W, n, m}. Therefore, the computational complexity of the MOEA/TS algorithm is O(mN3). As a reference algorithm, the computational complexity of the NSGA-III algorithm is O(mN2). The computational complexity of the MOEA/TS algorithm is an order of magnitude higher than that of the NSGA-III algorithm. This shows that the MOEA/TS algorithm is an expensive many-objective optimization algorithm.
It should be noted that although MOEA/TS algorithm has a higher computational complexity. But compared with the NSGA-III algorithm, the MOEA/TS algorithm also has greater advantages. In terms of convergence optimization, the NSGA-III algorithm adopts the traditional definition of Pareto domination. Obviously, the traditional definition can’t solve the problem of Pareto resistance. MOEA/TS algorithm uses the concept of “individual importance degree”. Individual importance degree can solve the problem of Pareto resistance. In terms of diversity optimization, the NSGA-III algorithm uses predefined reference points. The predefined reference points can’t solve the problem that the population can't be evenly distributed on the real PF in the high-dimensional objective space. MOEA/TS algorithm uses the repulsion field method. The repulsion field method can solve the problem that the population can’t be evenly distributed on the real PF in the high-dimensional objective space. In terms of algorithm architecture, the NSGA-III algorithm adopts the serial algorithm architecture. The serial algorithm architecture is difficult to coordinate the convergence optimization and diversity optimization of the algorithm. MOEA/TS algorithm adopts the concurrent algorithm architecture. The concurrent algorithm architecture can coordinate the convergence optimization and diversity optimization of the algorithm. In terms of operators, the NSGA-III algorithm uses the traditional binary crossover operator and polynomial mutation operator. The evolutionary ability of these two operators is weak. MOEA/TS algorithm uses feature extraction operator. Feature extraction operator has strong evolutionary ability. Therefore, the MOEA/TS algorithm has better performance than the NSGA-III algorithm. The comparison results support our conclusion. For the comparison results of these two algorithms, please browse Supplementary Information Document.
Experimental settings, configuration of experimental software and hardware.
The hardware and software configurations of the experiment are shown in Table 2 . Among them, PlatEMO 30 is a professional many-objective optimization experiment platform. The platform includes multiple test function sets and many-objective optimization algorithms.
The test functions used in the experiment include: DTLZ test function set (DTLZ1-7), MAF test function set (MAF1-6) and WFG test function set (WFG1-9). Literature 31 describes the characteristics of related test functions. The parameter settings of the related test functions are shown in Table 3 .
In order to verify the performance of MOEA/TS algorithm in the many-objective optimization field, this paper compares MOEA/TS algorithm with 7 advanced many-objective optimization algorithms. These 7 many-objective optimization algorithms include: VMEF 32 , BiGE-BEW 33 , MOEA/DG 34 , MOEA/D 35 , LSMaODE 36 , MaOEA/IT 23 and MaOEA/IGD 37 .
For all test cases, Wilcoxon rank sum test at 5% significance level 38 is used to compare the significance of the difference between the MOEA/TS algorithm and the comparison algorithms. The symbol “+” indicates that the comparison algorithms are significantly better than the MOEA/TS algorithm; the symbol “-“indicates that the comparison algorithms are significantly inferior to the MOEA/TS algorithm. The symbol “=” indicates that there is no significant difference between the MOEA/TS algorithm and the comparison algorithms.
In the aspect of performance evaluation, this paper uses inverted generational distance plus (IGD+) and hypervolume (HV) 39 to measure the performance of many-objective optimization algorithm. The smaller the IGD+ value that the algorithm obtains, the better the performance of the algorithm. The larger the HV value that the algorithm obtains, the better the performance of the algorithm.
In order to facilitate observation, we provide the normalized HV value of each algorithm relative to the best HV result. This normalization makes all the results lie in the range [0,1], and 1 represents the best value.
Considering the length of the paper, we only show the IGD+ values of different algorithms in the experiment chapter. For the HV values of different algorithms, please browse the Supplementary Information Document.
In terms of algorithm parameters, according to some existing parameter research results 13 , 40 , the feature factor \(\mu \) is set to 20 in this paper. According to the parameter sensitivity analysis, the number of high-quality solutions W is set to 9 in this paper. The parameter sensitivity analysis of W is detailed in the subsequent chapters.
The algorithm parameters of the 7 comparison algorithms are determined according to the best parameters provided by the corresponding literature.
Performance comparison under dtlz test function set.
In this paper, each algorithm is executed 30 times to get the average data as shown in Table 4 . As can be seen from Table 4 , MOEA/TS algorithm wins the first place in 15 test cases; BiGE-BEW algorithm wins the first place in 5 test cases; MOEA/D algorithm wins the first place in 15 test cases. In the 35 test cases, the number of MOEA/TS algorithm is significantly superior to VMEF algorithm, BiGE-BEW algorithm, MOEA/DG algorithm, MOEA/D algorithm, LSMaODE algorithm, MaOEA/IT algorithm and MaOEA/IGD algorithm is 21, 27, 25, 16, 32, 35 and 31, respectively. The number of MOEA/TS algorithm is significantly inferior to VMEF algorithm, BiGE-BEW algorithm, MOEA/DG algorithm, MOEA/D algorithm, LSMaODE algorithm, MaOEA/IT algorithm and MaOEA/IGD algorithm is 6, 5, 5, 15, 1, 0 and 0, respectively. Statistically, the number of MOEA/TS algorithm is similar to VMEF algorithm, BiGE-BEW algorithm, MOEA/DG algorithm, MOEA/D algorithm, LSMaODE algorithm, MaOEA/IT algorithm and MaOEA/IGD algorithm is 8, 3, 5, 4, 2, 0 and 4, respectively. Therefore, in the DTLZ test function set, MOEA/TS algorithm and MOEA/D algorithm have the best performance. The performance of VMEF algorithm, MOEA/DG algorithm, BiGE-BEW algorithm and LSMaODE algorithm decreases in turn. The performance of MaOEA/IGD algorithm and MaOEA/IT algorithm is similar and the worst.
Based on Table 4 , we further analyze the performance of these algorithms. In the DTLZ test function set, MOEA/TS algorithm performs poorly on DTLZ1, DTLZ5 and DTLZ6 test functions. The possible reasons are that the DTLZ1 test function has multiple local optima, and the DTLZ5 and DTLZ6 test functions have a narrow convergence curve. In the DTLZ1 test function, although the repulsion field method of the MOEA/TS algorithm makes the population widely distributed. However, its population distribution isn’t uniform and regular. The population distribution of some algorithms using predefined weight vectors is uniform and regular. In the DTLZ5 and DTLZ6 test functions, the coordination mechanism of MOEA/TS algorithm fails. The narrow convergence curve makes the population more concentrated, but the repulsion field method will disperse the population. The coordination mechanism is difficult to play a role.
The real Pareto front of DTLZ test function set is regular and the function complexity isn’t high. Therefore, algorithms with better diversity may be more popular. MOEA/D algorithm uses predefined weight vectors to maintain diversity and aggregation functions to maintain convergence. Therefore, it has good performance. VMEF algorithm uses different convergence ranking methods to deal with different test problems. Therefore, VMEF algorithm is good in convergence and poor in diversity. Based on the convergence measure and diversity measure, BiGE-BEW algorithm transforms the many-objective optimization problem into a two-objective optimization problem. In theory, the algorithm should perform well. However, there are defects in its convergence and diversity measurement formula. Finally, the experimental results of the algorithm aren’t as good as the expected results. MOEA/DG algorithm still uses the traditional dominance relationship to maintain the convergence of external archives. Therefore, MOEA/DG algorithm is poor in convergence and good in diversity. LSMaODE algorithm divides the population into two subpopulations and uses different strategies to optimize them. Because the real Pareto front of DTLZ test function set isn’t complex, the advantage of this multi-population algorithm architecture isn’t obvious. Therefore, compared with other algorithms, its performance is mediocre. MaOEA/IT algorithm optimizes convergence and diversity through two independent phases. However, the algorithm's performance is always poor because it doesn’t alleviate the contradiction between convergence and diversity. The reference Pareto front of MaOEA/IGD algorithm is poor. Therefore, the algorithm’s performance is always poor.
In this paper, each algorithm is executed 30 times to get the average data as shown in Table 5 . As can be seen from Table 5 , MOEA/TS algorithm wins the first place in 10 test cases; BiGE-BEW algorithm wins the first place in 8 test cases; MOEA/DG algorithm wins the first place in 2 test cases; MOEA/D algorithm wins the first place in 5 test cases; LSMaODE algorithm wins the first place in 5 test cases. In the 30 test cases, the number of MOEA/TS algorithm is significantly superior to VMEF algorithm, BiGE-BEW algorithm, MOEA/DG algorithm, MOEA/D algorithm, LSMaODE algorithm, MaOEA/IT algorithm and MaOEA/IGD algorithm is 22, 18, 25, 21, 20, 27 and 30, respectively. The number of MOEA/TS algorithm is significantly inferior to VMEF algorithm, BiGE-BEW algorithm, MOEA/DG algorithm, MOEA/D algorithm, LSMaODE algorithm, MaOEA/IT algorithm and MaOEA/IGD algorithm is 6, 11, 2, 5, 9, 1 and 0, respectively. Statistically, the number of MOEA/TS algorithm is similar to VMEF algorithm, BiGE-BEW algorithm, MOEA/DG algorithm, MOEA/D algorithm, LSMaODE algorithm, MaOEA/IT algorithm and MaOEA/IGD algorithm is 2, 1, 3, 4, 1, 2 and 0, respectively. Therefore, in the MAF test function set, MOEA/TS algorithm has the best performance. The performance of BiGE-BEW algorithm, LSMaODE algorithm, VMEF algorithm, MOEA/D algorithm, MOEA/DG algorithm and MaOEA/IT algorithm decreases in turn. The performance of MaOEA/IGD algorithm is the worst.
Based on Table 5 , we further analyze the performance of these algorithms. In the MAF test function set, MOEA/TS algorithm performs poorly on MAF2 and MAF3 test functions. The possible reasons are that the MAF2 test function greatly increases the difficulty of convergence on the basis of the DTLZ2 test function, and the MAF3 test function has a convex Pareto front and many local fronts. In the MAF2 test function, although the MOEA/TS algorithm can recognize the advantage and disadvantage of different individuals in the same front layer, the evolutionary efficiency of the MOEA/TS algorithm isn’t ideal. In other words, after the algorithm is finished, the population still has the large evolution potential in convergence. In the MAF3 test function, MOEA/TS algorithm can effectively deal with the convex Pareto front. However, MOEA/TS algorithm is difficult to deal with multiple local fronts because feature extraction operator of MOEA/TS algorithm is difficult to extract features of multiple local fronts.
MAF test function set is the variety of DTLZ test function set. It adds a lot of characteristics to the DTLZ test function set. For example, degenerate, convex, concave, partial, multimodal, deceptive, et al. Therefore, the MAF test function set is more difficult in terms of convergence and diversity. Based on the convergence measure and diversity measure, BiGE-BEW algorithm transforms the many-objective optimization problem into a two-objective optimization problem. Although there are some defects in its diversity and convergence measurement formula, BiGE-BEW algorithm shows good performance in convergence when dealing with more complex MaOPs. VMEF algorithm uses different convergence ranking methods to deal with different test problems. However, the complex Pareto fronts and diversified characteristics still pose a great challenge to VMEF algorithm. Therefore, the performance of VMEF algorithm is mediocre. MOEA/DG algorithm still uses the traditional dominance relationship to maintain the convergence of external archives. Therefore, MOEA/DG algorithm is poor in convergence. MOEA/D algorithm uses predefined weight vectors to maintain diversity and aggregation functions to maintain convergence. MOEA/D algorithm can easily deal with the DTLZ test function set. However, its performance isn’t ideal when dealing with more complex MAF test function set. Surprisingly, LSMaODE algorithm shows good performance. We speculate that the possible reason is that the real Pareto front of the MAF test function set is complex, and then the advantages of multi-population algorithm architecture can be reflected. MaOEA/IT algorithm optimizes convergence and diversity through two independent phases. However, the algorithm’s performance is always poor because it doesn’t alleviate the contradiction between convergence and diversity. The reference Pareto front of MaOEA/IGD algorithm is poor. Therefore, the algorithm’s performance is always poor.
In this paper, each algorithm is executed 30 times to get the average data as shown in Table 6 . As can be seen from Table 6 , MOEA/TS algorithm wins the first place in 27 test cases; VMEF algorithm wins the first place in 8 test cases; BiGE-BEW algorithm wins the first place in 6 test cases; MOEA/DG algorithm wins the first place in 1 test case; LSMaODE algorithm wins the first place in 3 test cases. In the 45 test cases, the number of MOEA/TS algorithm is significantly superior to VMEF algorithm, BiGE-BEW algorithm, MOEA/DG algorithm, MOEA/D algorithm, LSMaODE algorithm, MaOEA/IT algorithm and MaOEA/IGD algorithm is 26, 29, 42, 45, 39, 45 and 43, respectively. The number of MOEA/TS algorithm is significantly inferior to VMEF algorithm, BiGE-BEW algorithm, MOEA/DG algorithm, MOEA/D algorithm, LSMaODE algorithm, MaOEA/IT algorithm and MaOEA/IGD algorithm is 10, 9, 3, 0, 3, 0 and 0, respectively. Statistically, the number of MOEA/TS algorithm is similar to VMEF algorithm, BiGE-BEW algorithm, MOEA/DG algorithm, MOEA/D algorithm, LSMaODE algorithm, MaOEA/IT algorithm and MaOEA/IGD algorithm is 9, 7, 0, 0, 3, 0 and 2, respectively. Therefore, in the WFG test function set, MOEA/TS algorithm has the best performance. The performance of VMEF algorithm, BiGE-BEW algorithm, LSMaODE algorithm and MOEA/DG algorithm decreases in turn. The performance of MaOEA/IGD algorithm, MOEA/D algorithm and MaOEA/IT algorithm is similar and the worst.
Based on Table 6 , we further analyze the performance of these algorithms. MOEA/TS algorithm performs well in all WFG test functions. The possible reason is that the problem characteristics of the WFG test function set are bias, fraud and degradation. The WFG test function set is more difficult than the DTLZ test function set. However, the problem characteristics of the WFG test function set don’t include multiple local fronts (From the previous analysis, we know that MOEA/TS algorithm isn’t good at dealing with multiple local fronts.). MOEA/TS algorithm can deal with these problem characteristics. Therefore, MOEA/TS algorithm performs well in all WFG test functions. It should be noted that the WFG3 test function has a narrow convergence curve, but the performance of MOEA/TS algorithm is still the best. This is an interesting phenomenon. Because from the previous analysis, we know that MOEA/TS algorithm isn’t good at dealing with test functions with narrow convergence curves (such as DTLZ5 and DTLZ6 test functions). Based on the convergence difficulty of the WFG test function set, we speculate that the performance of the other 7 algorithms is worse, thus highlighting the performance of MOEA/TS algorithm.
Compared with the DTLZ test function set, the MAF test function set is more difficult in terms of convergence and diversity. VMEF algorithm uses different convergence ranking methods to deal with different test problems. This approach helps VMEF algorithm to deal with different problem characteristics. Therefore, the performance of VMEF algorithm is good. Based on the convergence measure and diversity measure, BiGE-BEW algorithm transforms the many-objective optimization problem into a two-objective optimization problem. Although there are some defects in its diversity and convergence measurement formula, BiGE-BEW algorithm shows good performance in convergence when dealing with more complex MaOPs. MOEA/DG algorithm still uses the traditional dominance relationship to maintain the convergence of external archives. Therefore, MOEA/DG algorithm is poor in convergence. MOEA/D algorithm uses predefined weight vectors to maintain diversity and aggregation functions to maintain convergence. This approach isn’t suitable for dealing with test functions with bias characteristic. Therefore, the performance of MOEA/D algorithm is the worst. LSMaODE algorithm divides the population into two subpopulations and uses different strategies to optimize them. Because most WFG test functions have bias characteristic, LSMaODE algorithm doesn’t consider the bias problem. Therefore, the performance of LSMaODE algorithm is mediocre. MaOEA/IT algorithm optimizes convergence and diversity through two independent phases. However, the algorithm’s performance is always poor because it doesn’t alleviate the contradiction between convergence and diversity. The reference Pareto front of MaOEA/IGD algorithm is poor. Therefore, the algorithm’s performance is always poor.
By synthesizing Tables 4 , 5 , 6 , we can obtain the data shown in Table 7 . As can be seen from Tables 4 , 5 , 6 , MOEA/TS algorithm wins the first place in 52 test cases; VMEF algorithm wins the first place in 8 test cases; BiGE-BEW algorithm wins the first place in 19 test cases; MOEA/DG algorithm wins the first place in 3 test cases; MOEA/D algorithm wins the first place in 20 test cases; LSMaODE algorithm wins first place in 8 test cases. As can be seen from Table 7 , in the 110 test cases, the number of MOEA/TS algorithm is significantly superior to VMEF algorithm, BiGE-BEW algorithm, MOEA/DG algorithm, MOEA/D algorithm, LSMaODE algorithm, MaOEA/IT algorithm and MaOEA/IGD algorithm is 69, 74, 92, 82, 91, 107 and 104, respectively. The number of MOEA/TS algorithm is significantly inferior to VMEF algorithm, BiGE-BEW algorithm, MOEA/DG algorithm, MOEA/D algorithm, LSMaODE algorithm, MaOEA/IT algorithm and MaOEA/IGD algorithm is 22, 25, 10, 20, 13, 1 and 0, respectively. Statistically, the number of MOEA/TS algorithm is similar to VMEF algorithm, BiGE-BEW algorithm, MOEA/DG algorithm, MOEA/D algorithm, LSMaODE algorithm, MaOEA/IT algorithm and MaOEA/IGD algorithm is 19, 11, 8, 8, 6, 2 and 6, respectively. Based on the above data, we can get the following conclusions: MOEA/TS algorithm has the best performance; the performance of BiGE-BEW algorithm, VMEF algorithm, MOEA/D algorithm, LSMaODE algorithm, MOEA/DG algorithm and MaOEA/IT algorithm decreases in turn. MaOEA/IGD algorithm has the worst performance.
In addition to the above conclusions, we can also observe 3 interesting phenomena:
(1) In the MAF test function set and WFG test function set, MOEA/TS algorithm has no competitors. However, in the DTLZ test function set, MOEA/TS algorithm and MOEA/D algorithm are competitors, and they have similar performance. This is because most DTLZ test functions have regular PF, while most MAF test functions and WFG test functions have more complex PF. It can be seen from Sect. " Introduction " that MOEA/D algorithm is suitable for MaOPs with regular PF. Therefore, in the DTLZ test function set, MOEA/D algorithm can compete with MOEA/TS algorithm. In the MAF test functions and WFG test functions, only MOEA/TS algorithm shows excellent performance.
(2) The performance of MOEA/TS algorithm is better on the test cases with 10 objectives, 15 objectives and 20 objectives. The performance of MOEA/TS algorithm is relatively ordinary on the test cases with 5 objectives and 8 objectives. This is because when the number of optimization objectives is small, most many-objective optimization algorithms perform well. Compared with other many-objective optimization algorithms, the advantages of MOEA/TS algorithm aren’t obvious. However, with the increase of the number of optimization objectives, the performance of other many-objective optimization algorithms becomes worse and worse. In contrast, the performance of MOEA/TS algorithm isn’t significantly affected. Therefore, compared with other many-objective optimization algorithms, MOEA/TS algorithm has obvious advantages. This shows that MOEA/TS algorithm is more suitable for solving MaOPs with more than 10 objectives.
(3) Without considering MOEA/TS algorithm, MOEA/D algorithm has the best performance in the DTLZ test function set. BiGE-BEW algorithm has the best performance in the MAF test function set. VMEF algorithm has the best performance in the WFG test function set. This shows that different many-objective optimization algorithms are suitable for different test function sets. However, MOEA/TS algorithm can show excellent performance on three test function sets. This indicates that MOEA/TS algorithm has strong universality and applicability.
In order to describe the distribution of solutions in the high-dimensional objective space more intuitively, this paper draws the distribution diagram of solutions in the objective space. Considering the length of the paper, it is unrealistic to show the distribution diagrams of all test functions. Therefore, this section only shows the distribution diagrams of 3 representative test cases. These 3 test cases are DTLZ2 test case with 20 objectives, MAF1 test case with 15 objectives and WFG3 test case with 10 objectives, respectively.
Figure 10 shows the distribution diagrams of each algorithm on DTLZ2 test case with 20 objectives. It can be seen from Fig. 10 that distribution diagrams of MOEA/TS algorithm, BiGE-BEW algorithm, MOEA/DG algorithm and MOEA/D algorithm are similar, which indicates that these 4 algorithms are excellent in convergence and diversity; VMEF algorithm and LSMaODE algorithm are good in diversity, but poor in convergence; MaOEA/IT algorithm and MaOEA/IGD algorithm are very poor in convergence and diversity.
Distribution diagrams of each algorithm on DTLZ2 test case with 20 objectives.
Figure 11 shows the distribution diagrams of each algorithm on MAF1 test case with 15 objectives. It can be seen from Fig. 11 that MOEA/TS algorithm and VMEF algorithm are good in convergence, but poor in diversity; BiGE-BEW algorithm and LSMaODE algorithm are good in diversity, but poor in convergence. MOEA/DG algorithm, MOEA/D algorithm, MaOEA/IT algorithm and MaOEA/IGD algorithm are very bad in convergence and diversity.
Distribution diagrams of each algorithm on MAF1 test case with 15 objectives.
Figure 12 shows the distribution diagrams of each algorithm on WFG3 test case with 10 objectives. It can be seen from Fig. 12 that MOEA/TS algorithm has the best convergence and diversity; LSMaODE algorithm is also excellent, only slightly worse than MOEA/TS algorithm in terms of diversity; BiGE-BEW algorithm and MOEA/DG algorithm are good in diversity, but poor in convergence. VMEF algorithm is good in convergence, but poor in diversity. MOEA/D algorithm, MaOEA/IT algorithm and MaOEA/IGD algorithm are very bad in convergence and diversity.
Distribution diagrams of each algorithm on WFG3 test case with 10 objectives.
This section takes DTLZ2 test case with 20 objectives, MAF1 test case with 15 objectives and WFG3 test case with 10 objectives as examples to display the evolution curves of 8 algorithms (as shown in Figs. 13 , 14 , 15 ).
Evolution curve of each algorithm on DTLZ2 test case with 20 objectives.
In Figure 13 , in terms of the final IGD+ value of the algorithm, MOEA/TS algorithm has the smallest IGD+ value, while the IGD+ values of MOEA/DG algorithm, BiGE-BEW algorithm, MOEA/D algorithm, LSMaODE algorithm, VMEF algorithm and MaOEA/IGD algorithm successively increase, and MaOEA/IT algorithm has the largest IGD+ value. This shows that MOEA/TS algorithm has the best convergence and diversity within the specified number of iterations. In terms of the evolution of the algorithm, the final IGD+ values of all algorithms are smaller than the initial IGD+ values. This shows that all algorithms have strong evolution ability, especially MOEA/TS algorithm has the strongest evolution ability. In terms of algorithm fluctuation, MaOEA/IT algorithm fluctuates greatly. This shows that MaOEA/IT algorithm isn’t stable. Based on the above analysis, we believe that MOEA/TS algorithm has the best comprehensive performance on DTLZ2 test case with 20 objectives, and is suitable for solving DTLZ2 test problem with 20 objectives.
In Figure 14 , in terms of the final IGD+ value of the algorithm, MOEA/TS algorithm has the smallest IGD+ value, while the IGD+ values of BiGE-BEW algorithm, VMEF algorithm, LSMaODE algorithm, MaOEA/IGD algorithm, MOEA/D algorithm and MOEA/DG algorithm successively increase, and MaOEA/IT algorithm has the largest IGD+ value. This shows that MOEA/TS algorithm has the best convergence and diversity within the specified number of iterations. In terms of the evolution of the algorithm, the final IGD+ values of all algorithms are smaller than the initial IGD+ values. This shows that all algorithms have strong evolution ability, especially MOEA/TS algorithm has the strongest evolution ability. In terms of algorithm fluctuation, MaOEA/IT algorithm fluctuates greatly. This shows that MaOEA/IT algorithm isn’t stable. Based on the above analysis, we believe that MOEA/TS algorithm has the best comprehensive performance on MAF1 test case with 15 objectives, and is suitable for solving MAF1 test problem with 15 objectives.
Evolution curve of each algorithm on MAF1 test case with 15 objectives.
In Fig. 15 , in terms of the final IGD+ value of the algorithm, MOEA/TS algorithm has the smallest IGD+ value, while the IGD+ values of LSMaODE algorithm, MOEA/DG algorithm, VMEF algorithm, BiGE-BEW algorithm, MaOEA/IGD algorithm and MOEA/D algorithm successively increase, and MaOEA/IT algorithm has the largest IGD+ value. This shows that MOEA/TS algorithm has the best convergence and diversity within the specified number of iterations. In terms of the evolution of the algorithm, the final IGD+ values of the MaOEA/IT algorithm, VMEF algorithm, MaOEA/IGD algorithm, BiGE-BEW algorithm and VMEF algorithm are all greater than the initial IGD+ values. This shows that the performance of these 5 algorithms deteriorates during evolution, and they aren’t suitable for dealing with WFG3 test problem with 10 objectives. The initial IGD+ value of MOEA/DG algorithm is close to the final IGD+ value, and the IGD+ value of MOEA/DG algorithm fluctuates little during the evolution. This shows that MOEA/DG algorithm is insensitive to evolution. Only the final IGD+ values of LSMaODE algorithm and MOEA/TS algorithm are less than the initial IGD+ values. This shows that LSMaODE algorithm and MOEA/TS algorithm have strong evolution ability, especially MOEA/TS algorithm has the strongest evolution ability. In terms of algorithm fluctuation, MOEA/D algorithm, MaOEA/IT algorithm and MaOEA/IGD algorithm have greater fluctuation. This shows that these 3 algorithms aren’t stable. Based on the above analysis, we believe that MOEA/TS algorithm has the best comprehensive performance on WFG3 test case with 10 objectives, and is suitable for solving WFG3 test problem with 10 objectives.
Evolution curve of each algorithm on WFG3 test case with 10 objectives.
In addition, we can also observe an interesting phenomenon from Fig. 13 to Fig. 15 : the IGD+ values of some algorithms sometimes increase significantly with the increase of iterations. That is, the performance of some algorithms sometimes deteriorates seriously with the increase of iterations. The reasons for this phenomenon may include three aspects: (1) The algorithm doesn’t adopt the elite preservation strategy. Some high-quality solutions may gradually disappear; (2) Due to the complexity of the optimization problems, the evolutionary direction of the population may be misled by some pseudo-elite individuals; (3) The convergence optimization and diversity optimization of the algorithm aren’t coordinated. The optimization of convergence may affect the optimization of diversity or the optimization of diversity may affect the optimization of convergence. It can be seen from the pseudo-code of the algorithm in Section 3.5 that the MOEA/TS algorithm proposed in this paper considers the above three aspects. Therefore, MOEA/TS algorithm can effectively alleviate this phenomenon.
In order to verify the effectiveness of the innovative parts, 4 variants are designed in this section. As follows:
MOEA/TS-1 algorithm: The feature extraction operator in MOEA/TS algorithm is changed to the binary crossover operator and polynomial mutation operator;
MOEA/TS-2 algorithm: The repulsion field method in MOEA/TS algorithm is removed;
MOEA/TS-3 algorithm: The concurrent architecture in MOEA/TS algorithm is changed to serial architecture;
MOEA/TS-4 algorithm: The individual importance degree in MOEA/TS algorithm is removed.
This paper takes WFG test function set (45 test cases) as samples, and then verifies the performance of 5 algorithms. In this paper, 5 algorithms are executed 30 times to get the average data as shown in Table 8 . As can be seen from Table 8 , MOEA/TS algorithm wins the first place in 24 test cases; MOEA/TS-1 algorithm wins the first place in 13 test cases; MOEA/TS-2 algorithm wins the first place in 7 test cases; MOEA/TS-3 algorithm wins the first place in 1 test case. In the 45 test cases, the number of MOEA/TS algorithm is significantly superior to MOEA/TS-1 algorithm, MOEA/TS-2 algorithm, MOEA/TS-3 algorithm and MOEA/TS-4 algorithm is 21, 30, 40 and 45, respectively. The number of MOEA/TS algorithm is significantly inferior to MOEA/TS-1 algorithm, MOEA/TS-2 algorithm, MOEA/TS-3 algorithm and MOEA/TS-4 algorithm is 11, 6, 0 and 0, respectively. Statistically, the number of MOEA/TS algorithm is similar to MOEA/TS-1 algorithm, MOEA/TS-2 algorithm, MOEA/TS-3 algorithm and MOEA/TS-4 algorithm is 13, 9, 5 and 0, respectively. The average ranking of MOEA/TS algorithm is about 1.64; the average ranking of MOEA/TS-1 algorithm is about 2.02; the average ranking of MOEA/TS-2 algorithm is about 2.62; the average ranking of MOEA/TS-3 algorithm is about 3.71; the average ranking of MOEA/TS-4 algorithm is 5.
Therefore, we think that the 4 innovative parts of MOEA/TS algorithm are necessary and indispensable. The lack of any innovative parts will seriously affect the performance of MOEA/TS algorithm. This shows that our innovations are effective. In addition, based on the above data, we can also find that “individual importance degree” has the greatest influence on the algorithm; the algorithm architecture ranks second; the repulsion field method ranks third; the feature extraction operator ranks fourth.
In the feature extraction operator, we select W high-quality solutions. To prove the effectiveness of this selection approach over random selection, the ablation experiment will be performed in this sect. " Introduction " variant is designed in this section. As follows:
MOEA/TS-5 algorithm: W solutions are randomly selected in the feature extraction operator.
This paper takes WFG test function set (45 test cases) as samples, and then verifies the performance of 2 algorithms. In this paper, 2 algorithms are executed 30 times to get the average data as shown in Table 9 . As can be seen from Table 9 , MOEA/TS algorithm wins the first place in 45 test cases. In the 45 test cases, the number of MOEA/TS algorithm is significantly superior to MOEA/TS-5 algorithm is 42. The number of MOEA/TS algorithm is significantly inferior to MOEA/TS-5 algorithm is 0. Statistically, the number of MOEA/TS algorithm is similar to MOEA/TS-5 algorithm is 3. Therefore, we believe that the performance of MOEA/TS algorithm is better than MOEA/TS-5 algorithm in the WFG test function set. It proves that the selection approach that we use is better than random selection in the feature extraction operator.
In addition, the performance of MOEA/TS-5 algorithm isn’t as good as that of MOEA/TS-1 algorithm. It means that the performance of the feature extraction operator based on random selection is even worse than that of some classical operators. The possible reason is that the randomly selected solution set will cause the feature extraction operator to extract many bad features. These bad features hinder individual evolution, which makes the convergence maintenance state and diversity maintenance state of MOEA/TS algorithm fail for a long time, and only the coordination state can play some role. The architecture of the MOEA/TS algorithm is undermined by some bad features.
The algorithm parameters analyzed in this paper are mainly the number of high-quality solutions W, threshold value T, standard deviation std. Due to the high complexity of the WFG3 test case with 10 objectives, it is difficult for the population of each algorithm to cover the real Pareto front, so this paper considers the WFG3 test case with 10 objectives as the main function of parameter analysis.
The initial value and value range of each parameter are shown in Table 10 .
As shown in Fig. 16 , when \(W<9\) , the IGD + value of the algorithm decreases significantly with the increase of W . It means that when \(W<9\) , the performance of the feature extraction operator is greatly improved with the increase of W . This is because the features extracted by the feature extraction operator are closer to the ideal situation. When \(W=9\) , the IGD + value of the algorithm is minimum. This shows that when \(W=9\) , the feature extraction operator performs best. When \(W>9\) , the IGD + value of the algorithm increases slowly. It means that when \(W>9\) , the performance of the feature extraction operator deteriorates gradually with the increase of W . This is because some features are over-extracted by feature extraction operators. Therefore, for WFG3 test case with 10 objectives, \(W=9\) is the best parameter selection.
The corresponding relationship between IGD + value and W.
As shown in Fig. 17 , when \(T<5\%\) , the IGD + value of the algorithm decreases significantly with the increase of T . This is because if the threshold value T is too small, the algorithm will remain in the same state for a long time, and it is difficult to be adjusted to other states. Convergence and diversity of algorithm will also be difficult to balance. This situation will be improved with the increase of T . When \(T=5\%\) , the IGD + value of the algorithm is minimum. This shows that when \(T=5\%\) , the algorithm has the best performance. When \(T>5\%\) , the IGD + value of the algorithm increases gradually with the increase of T . This is because if the threshold value T is too large, the algorithm’s state will be adjusted frequently. Even if the population isn’t stable in one state (convergence, diversity, coordination), the algorithm will also be adjusted to other states. This isn’t conducive to improving the convergence and the diversity of the algorithm. The efficiency of the algorithm will also be affected. Therefore, for WFG3 test case with 10 objectives, \(T=5\%\) is the best parameter selection.
The corresponding relationship between IGD + value and T.
As shown in Fig. 18 , when \(std<0.7\) , the IGD + value of the algorithm decreases significantly with the increase of std . This is because if std is too small, the results of Gaussian sampling are too concentrated in the middle region, and the randomness of the sampling vector is weak, which isn’t conducive to the use of features and generation of diversified feature solutions. When \(std=0.7\) , the IGD + value of the algorithm is minimum. This shows that when \(std=0.7\) , the feature extraction operator performs best. When \(std>0.7\) , the IGD + value of the algorithm increases significantly with the increase of std . This is because if the std is too large, the result of Gaussian sampling is too scattered, the randomness of the sampling vector is strong, some components are easy to exceed the upper bound or lower bound, and some features are easy to be eliminated by the repair operator. Therefore, for WFG3 test case with 10 objectives, \(std=0.7\) is the best parameter selection.
The corresponding relationship between IGD + value and std.
Based on the above analysis of algorithm parameters, we think \(W=9, T=5\%, std=0.7\) are the best parameter combinations in WFG3 test case with 10 objectives. Further, we test the performance of the above parameter combinations in more test cases. The experimental results show that the above parameter combinations perform well in most test cases. Therefore, this paper sets the number of high-quality solutions \(W\) , the threshold value \(T\) and the standard deviation \(std\) to 9, 5% and 0.7, respectively.
This section mainly explores the performance of MOEA/TS algorithm in practical problems. The practical problem selected in this section is the industrial internet optimization problem based on the blockchain provided in reference 40 .
The industrial internet can support effective control of the physical world through a large amount of industrial data, but data security has always been a challenge due to various interconnections and accesses. Blockchain technology supports the security and privacy protection of industrial internet data with its trusted and reliable security mechanism. Fragmentation technology can help improve the overall throughput and scalability of the blockchain network. However, due to the uneven distribution of malicious nodes, the effectiveness of fragmentation is still challenging. In addition, the conflict between multiple industrial network indicators is also a problem we have to consider. Therefore, the industrial internet optimization problem based on blockchain is an important research problem.
In this section, the industrial internet optimization problem based on blockchain has the following 4 optimization objectives:
(1) Minimizing the shard invalidation probability (SIP);
(2) Minimizing the transmission delay (TD);
(3) Maximizing the throughput (TP);
(4) Minimizing the load of Malicious Nodes (LMN).
The research background of the industrial internet based on blockchain and the calculation formulas of these 4 objectives are detailed in reference 40 .
In this section, we set the population size to 220, the number of iterations to 300, and the number of function evaluations to 66000. We still use inverted generational distance plus (IGD+) to measure the performance of many-objective optimization algorithms. However, the real PF of the practical problem is unknown. Therefore, we run these algorithms many times to obtain the different non-dominated solution sets. The non-dominated union set of the different non-dominated solution sets is considered as the real PF. The relevant parameters of these algorithms are shown in Section 4.1.
In this section, each algorithm is executed 30 times to get the data as shown in Table 11 . As can be seen from Table 11 , MOEA/TS algorithm has absolute advantages. The performance of BiGE-BEW algorithm and MOEA/DG algorithm is good and similar. The performance of VMEF algorithm and MOEA/D algorithm in practical problems is obviously not as good as that in benchmark test functions. This is because the real PF of the practical problem is more complex. The performance of LSMaODE algorithm is close to that of MOEA/D algorithm. The performance of MaOEA/IT algorithm and MaOEA/IGD algorithm is the worst. Based on the above observations and analysis, we believe that MOEA/TS algorithm still has excellent performance and strong applicability in practical problems.
Considering that the solutions obtained by the many-objective optimization algorithms are the population, it is unrealistic to compare different network indicators of different algorithms intuitively. However, in practical applications, we only need to make choices according to the specific needs or preferences of users or enterprises. In this section, we first select the individuals with the largest throughput in each algorithm, and then compare the MOEA/TS algorithm with other algorithms on the basis of ensuring the maximum throughput. The network indicators obtained by these 8 algorithms are shown in Table 12 . As can be seen from Table 12 , in terms of SIP and TP, MOEA/TS algorithm has the best performance; In terms of TD, MOEA/TS algorithm ranks second; In terms of LMN, MOEA/TS algorithm ranks third. Therefore, we believe that the MOEA/TS algorithm has the best comprehensive performance in the industrial internet optimization problem based on blockchain, and various network indicators are at the forefront.
Based on the experimental analysis from Section 4.2 to Section 4.8, we can obtain the following conclusions:
(1) In the benchmark test cases, MOEA/TS algorithm is superior to the other 7 advanced many-objective optimization algorithms.
(2) MOEA/TS algorithm is more suitable for dealing with the MaOPs with more than 10 objectives.
(3) MOEA/TS algorithm can show excellent performance in different test function sets, and has strong universality and applicability.
(4) MOEA/TS algorithm has the best convergence and diversity, the strongest evolution ability and the fastest convergence speed.
(5) The 4 innovative parts of MOEA/TS algorithm are necessary and indispensable. The lack of any innovative parts will seriously affect the performance of MOEA/TS algorithm.
(6) MOEA/TS algorithm still has excellent performance and strong applicability in practical problems.
Aiming at some difficulties in the many-objective optimization field, this paper proposes a many-objective evolutionary algorithm based on three states (MOEA/TS). Firstly, a feature extraction operator is proposed. The feature extraction operator is a feature extractor, which can extract the features of the high-quality solution set, and then assist the evolution of the current individual. Secondly, in terms of convergence maintenance, this paper doesn’t consider using domination relaxation technique, because the current domination relaxation technique still faces some problems. Based on Pareto front layer, this paper proposes the concept of “individual importance degree”. The importance degree of an individual can reflect the importance of the individual in the same Pareto front layer, so as to further distinguish the advantages and disadvantages of different individuals in the same front layer, and effectively solve the phenomenon of Pareto resistance. Then, in terms of diversity maintenance, this paper considers maintaining the diversity of the population in the objective space by repulsion field, so that the population can be evenly distributed on the real PF. Finally, a new concurrent algorithm framework is designed. In the framework, the algorithm is divided into three states, namely, convergence maintenance state, diversity maintenance state and coordination state. Each state focuses on a specific task. That is, the convergence maintenance state is responsible for improving the convergence of population; Diversity maintenance state is responsible for improving the diversity of population; the coordination state is responsible for coordinating the contradiction between diversity and convergence. The population can freely switch among these three states according to its own evolution. The experimental results show that MOEA/TS algorithm is superior to the other 7 advanced many-objective optimization algorithms. In addition, the effectiveness of the innovation parts is further verified.
However, MOEA/TS algorithm also has obvious defects: MOEA/TS algorithm isn’t good at dealing with test problems with narrow convergence curves or multiple local fronts. Therefore, in the future, we will further improve MOEA/TS algorithm, so that MOEA/TS algorithm can deal with test problems with narrow convergence curve or multiple local fronts. In addition, constrained MOPs and high-dimensional MOPs are also the focus of our future research.
All data generated or analysed during this study are included in this published article.
Lin, H. F. & Tang, C. P. Analysis and optimization of urban public transport lines based on multiobjective adaptive particle swarm optimization. IEEE Trans. Intell. Transp. Syst. 23 , 16786–16798. https://doi.org/10.1109/TITS.2021.3086808 (2022).
Article Google Scholar
Qin, X., Fang, Z. H. & Zhang, Z. X. Multi-objective optimization for production scheduling ofprecast components considering resource constraints. Comput. Int. Manuf. Syst. 27 , 2248–2259. https://doi.org/10.13196/j.cims.2021.08.008 (2021).
Saric, F., Begusic, S., Mercep, A. & Kostanjcar, Z. Statistical arbitrage portfolio construction based on preference relations. Expert Syst. Appl. 238 , 1–12. https://doi.org/10.1016/j.eswa.2023.121906 (2023).
Chen, Y., Zhong, J., Feng, L. & Zhang, J. An adaptive archive-based evolutionary framework for many-task optimization. IEEE Trans. Emerg. Top. Comput. Intell. 4 , 369–384. https://doi.org/10.1109/TETCI.2019.2916051 (2020).
Article CAS Google Scholar
Pradhan, D., Wang, S., Ali, S., Yue, T. & Liaaen, M. CBGA-ES+: A cluster-based genetic algorithm with non-dominated elitist selection for supporting multi-objective test optimization. IEEE Trans. Softw. Eng. 47 , 86–107. https://doi.org/10.1109/TETCI.2019.2916051 (2021).
Bian, H. L., Tian, J., Yu, J. L. & Yu, H. Bayesian co-evolutionary optimization based entropy search for high-dimensional many-objective optimization. Knowl. Based Syst. 274 , 1–13. https://doi.org/10.1016/j.knosys.2023.110630 (2023).
Li, W., Chen, Y. T., Dong, Y. H. & Huang, Y. A solution potential-based adaptation reference vector evolutionary algorithm for many-objective optimization. Swarm Evol. Comput. 84 , 1–15. https://doi.org/10.1016/j.swevo.2023.101451 (2023).
Wang, Y. J., Gao, P. & Chen, Y. An improved farmland fertility algorithm for many-objective optimization problems. Sci. Rep. 12 , 1–24. https://doi.org/10.1038/s41598-022-06329-x (2022).
Khurana, D., Yadav, A. & Sadollah, A. A non-dominated sorting based multi-objective neural network algorithm. MethodsX 10 , 1–16. https://doi.org/10.1016/j.mex.2023.102152 (2023).
Huang, H. J., Zheng, B. F., Wei, X. X., Zhou, Y. Q. & Zhang, Y. D. NSCSO: A novel multi-objective non-dominated sorting chicken swarm optimization algorithm. Sci. Rep. 14 , 1–38. https://doi.org/10.1038/s41598-024-54991-0 (2024).
Li, M. Q., Yang, S. X. & Liu, X. H. Bi-goal evolution for many-objective optimization problems. Artif. Intell. 228 , 45–65. https://doi.org/10.1016/j.artint.2015.06.007 (2015).
Article MathSciNet Google Scholar
Wei, L. S. & Li, E. C. A many-objective evolutionary algorithm with population preprocessing and projection distance-assisted elimination mechanism. J. Comput. Design Eng. 10 , 1988–2018. https://doi.org/10.1093/jcde/qwad088 (2023).
Sun, Y. F., Bian, K., Liu, Z., Sun, X. & Yao, R. X. Adaptive strategies based on differential evolutionary algorithm for many-objective optimization. Discret. Dyn. Nat. Soc. 2021 , 1–17. https://doi.org/10.1155/2021/2491796 (2021).
Zhou, S. Q., Dai, Y. R. & Chen, Z. H. Dominance relation selection and angle-based distribution evaluation for many-objective evolutionary algorithm. Swarm Evol. Comput. 86 , 1–19. https://doi.org/10.1016/j.swevo.2024.101515 (2024).
Wang, X. W., Xie, Z. H., Zhou, X. & Gu, X. S. A two-stage adaptive reference direction guided evolutionary algorithm with modified dominance relation for many-objective optimization. Swarm Evol. Comput. 78 , 1–14. https://doi.org/10.1016/j.swevo.2023.101272 (2023).
Zhang, W., Liu, J. C., Liu, J. H., Liu, Y. C. & Tan, S. B. A dual distance dominance based evolutionary algorithm with selection-replacement operator for many-objective optimization. Expert Syst. Appl. 237 , 1–25. https://doi.org/10.1016/j.eswa.2023.121244 (2023).
Shang, K. & Ishibuchi, H. A new hypervolume-based evolutionary algorithm for many-objective optimization. IEEE Trans. Evol. Comput. 24 , 839–852. https://doi.org/10.1109/TEVC.2020.2964705 (2020).
Zhang, W., Liu, J. C., Liu, J. H., Liu, Y. C. & Wang, H. H. A many-objective evolutionary algorithm based on novel fitness estimation and grouping layering. Neural Comput. Appl. 35 , 24283–24314. https://doi.org/10.1007/s00521-023-08950-x (2023).
Nan, Y., Shang, K., Ishibuchi, H. & He, L. J. A Two-stage Hypervolume Contribution Approximation Method Based on R2 Indicator. In: 2021 IEEE Congress on Evolutionary Computation (IEEE CEC 2021) . 2468-2475. https://doi.org/10.1109/CEC45853.2021.9504726 (2021).
Wu, M., Li, K., Kwong, S. & Zhang, Q. Evolutionary many-objective optimization based on adversarial decomposition. IEEE Trans. Cybern. 50 , 753–764. https://doi.org/10.1109/TCYB.2018.2872803 (2020).
Article PubMed Google Scholar
Fan, M. W. et al. Improved multi-objective differential evolution algorithm based on a decomposition strategy for multi-objective optimization problems. Sci. Rep. 12 , 1–14. https://doi.org/10.1038/s41598-022-25440-7 (2022).
Peng, F. A., Lv, L., Chen, W. R. & Wang, J. A projection-based evolutionary algorithm for multi-objective and many-objective optimization. Processes 11 , 1–22. https://doi.org/10.3390/pr11051564 (2023).
Sun, Y., Xue, B., Zhang, M. & Yen, G. G. A new two-stage evolutionary algorithm for many-objective optimization. IEEE Trans. Evol. Comput. 23 , 748–761. https://doi.org/10.1109/TEVC.2018.2882166 (2019).
Yang, Q. T., Zhan, Z. H., Kwong, S. & Zhang, J. Multiple populations for multiple objectives framework with bias sorting for many-objective optimization. IEEE Trans. Evol. Comput. 27 , 1340–1354. https://doi.org/10.1109/TEVC.2022.3212058 (2023).
Liu, S. C. et al. Many-objective job-shop scheduling: A multiple populations for multiple objectives-based genetic algorithm approach. IEEE Trans. Cybern. 53 , 1460–1474. https://doi.org/10.1109/TCYB.2021.3102642 (2023).
Tian, Y., He, C., Cheng, R. & Zhang, X. A multistage evolutionary algorithm for better diversity preservation in multiobjective optimization. IEEE Trans. Syst. Man Cybern. Syst. 51 , 5880–5894. https://doi.org/10.1109/TSMC.2019.2956288 (2021).
Sun, C. H., Wang, Y. H., Wan, P. & Du, Y. A cooperative spectrum sensing algorithm based on principal component analysis and K-medoids clustering. In: 2018 33rd Youth Academic Annual Conference of Chinese Association of Automation (YAC). 835-839. https://doi.org/10.1109/YAC.2018.8406487 (2018).
Osinsky, A., Bychkov, R., Trefilov, M., Lyashev, V. & Ivanov, A. Regularization for cholesky decomposition in massive MIMO detection. IEEE Wirel. Commun. Lett. 12 , 1603–1607. https://doi.org/10.1109/LWC.2023.3284349 (2023).
Gu, Q. H., Gao, S., Li, X. X., Xiong, N. N. & Liu, R. R. An adaptive adjacent maximum distance crossover operator for multi-objective algorithms. Soft Comput. 27 , 7419–7438. https://doi.org/10.1007/s00500-023-07978-4 (2023).
Tian, Y., Cheng, R., Zhang, X. Y. & Jin, Y. PlatEMO: A MATLAB platform for evolutionary multi-objective optimization [educational forum]. IEEE Comput. Intell. Mag. 12 , 73–87. https://doi.org/10.1109/MCI.2017.2742868 (2017).
He, Z., Yen, G. G. & Zhang, J. Fuzzy-based Pareto optimality for many-objective evolutionary algorithms. IEEE Trans. Evol. Comput. 18 , 269–285. https://doi.org/10.1109/TEVC.2013.2258025 (2014).
Qiu, W. et al. Ensemble many-objective optimization algorithm based on voting mechanism. IEEE Trans. Syst. Man Cybern. Syst. 52 , 1716–1730. https://doi.org/10.1109/TSMC.2020.3034180 (2020).
Wang, J. & Chen, H. A Weight Vector Bi-Objective Evolutionary Algorithm with Bi-criterion Evolution for Many-Objective Optimization. In: 2022 IEEE 4th International Conference on Power, Intelligent Computing and Systems (ICPICS). 273-279. https://doi.org/10.1109/ICPICS55264.2022.9873807 (2022).
He, X. & Dai, C. An Improvement Evolutionary Algorithm Based on Decomposition and Grid-based Pareto Dominance for Many-objective Optimization. In: 2022 Global Conference on Robotics, Artificial Intelligence and Information Technology (GCRAIT). 145-149. https://doi.org/10.1109/GCRAIT55928.2022.00039 (2022).
Zhang, Q., Liu, W. & Li, H. The performance of a new version of MOEA/D on CEC09 unconstrained MOP test instances. In: 2009 IEEE Congress on Evolutionary Computation (CEC). 203–208. https://doi.org/10.1109/CEC . 2009.4982949 (2009).
Zhang, K., Shen, C. & Yen, G. G. Multipopulation-based differential evolution for large-scale many-objective optimization. IEEE Trans. Cybern. 53 , 7596–7608. https://doi.org/10.1109/TCYB.2022.3178929 (2023).
Sun, Y., Yen, G. G. & Yi, Z. IGD indicator-based evolutionary algorithm for many-objective optimization problems. IEEE Trans. Evol. Comput. 23 , 173–187. https://doi.org/10.1109/TEVC.2018.2791283 (2019).
Wilcoxon, F. Individual comparisons by ranking methods. Breakthr. Stat. 1 , 196–202. https://doi.org/10.1007/978-1-4612-4380-9_16 (1992).
Ishibuchi, H., Imada, R., Masuyama, N. & Nojima, Y. Comparison of Hypervolume, IGD and IGD+ from the Viewpoint of Optimal Distributions of Solutions. In: 2019 International Conference on Evolutionary Multi-Criterion Optimization (EMO). 332–345. https://doi.org/10.1007/978-3-030-12598-1_27 (2019).
Cai, X. J. et al. A sharding scheme-based many-objective optimization algorithm for enhancing security in blockchain-enabled industrial internet of things. IEEE Trans. Ind. Inform. 17 , 7650–7658. https://doi.org/10.1109/TII.2021.3051607 (2021).
Download references
This work was supported in part by the National Natural Science Foundation of China under Grant 62362026 and 62162021; in part by the specific research fund of The Innovation Platform for Academicians of Hainan Province under grant YSPTZX202314; in part by the Key Project of Hainan Province under Grant ZDYF2023GXJS158.
National Natural Science Foundation of China, 62362026, 62362026, Specific research fund of The Innovation Platform for Academicians of Hainan Province, YSPTZX202314, YSPTZX202314, Key Project of Hainan Province, ZDYF2023GXJS158, ZDYF2023GXJS158.
Authors and affiliations.
School of Information and Communication Engineering, Hainan University, Haikou, 570228, China
School of Computer Science and Technology, Hainan University, Haikou, 570228, China
Huanhuan Yu, Hansheng Fei, Xiangdang Huang & Qiuling Yang
Academic Affairs Office, Shandong Agricultural University, Taian, 271018, China
Huijie Zhang
Innovation Platform for Academicians of Hainan Province, Hainan University, Haikou, 570228, China
Jiale Zhao, Huanhuan Yu, Hansheng Fei, Xiangdang Huang & Qiuling Yang
You can also search for this author in PubMed Google Scholar
Conceptualization, J.L.Z., H.J.Z. and X.D.H.; Methodology, J.L.Z., H.J.Z. and X.D.H.; Software, J.L.Z., H.J.Z., H.H.Y. and H.S.F.; Validation, H.H.Y. and H.S.F.; Formal analysis, J.L.Z. and H.J.Z.; Investigation, H.H.Y. and H.S.F.; Resources, Q.L.Y.; Data curation, H.H.Y. and H.S.F.; Writing-original draft, J.L.Z.; Writing-review&editing, J.L.Z., H.J.Z., H.H.Y. and H.S.F.; Visualization, J.L.Z. and H.J.Z.; Supervision, H.H.Y. and H.S.F.; Project administration, J.L.Z. and Q.L.Y.; Funding acquisition, Q.L.Y.; All authors have read and agreed to the published version of the manuscript.
Correspondence to Qiuling Yang .
Competing interests.
The authors declare no competing interests.
Publisher's note.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information., rights and permissions.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ .
Reprints and permissions
Cite this article.
Zhao, J., Zhang, H., Yu, H. et al. A many-objective evolutionary algorithm based on three states for solving many-objective optimization problem. Sci Rep 14 , 19140 (2024). https://doi.org/10.1038/s41598-024-70145-8
Download citation
Received : 26 January 2024
Accepted : 13 August 2024
Published : 19 August 2024
DOI : https://doi.org/10.1038/s41598-024-70145-8
Anyone you share the following link with will be able to read this content:
Sorry, a shareable link is not currently available for this article.
Provided by the Springer Nature SharedIt content-sharing initiative
By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.
Sign up for the Nature Briefing: AI and Robotics newsletter — what matters in AI and robotics research, free to your inbox weekly.
Veterans Data Hub Codebook
Teague Ruder , Megan S. Schuler
Tool Published Aug 21, 2024
This codebook was compiled as part of the RAND Veterans Policy Research Institute's effort to develop a repository of datasets that assess veteran status. This codebook for the American Community Survey Public Use Microdata Sample (ACS-PUMS) presents an overview of the data, how to access them, which variable(s) assess veteran status, other veteran-related questions, and additional resources for using the ACS-PUMS. The authors of this codebook cite directly from the original documentation of the ACS-PUMS.
Oct 24, 2023
Aug 21, 2024
Jul 15, 2021
Funding for this publication was made possible by a generous gift from Daniel J. Epstein through the Epstein Family Foundation. The research was conducted by the RAND Epstein Family Veterans Policy Research Institute within RAND Education and Labor .
This publication is part of the RAND tool series. RAND tools include models, databases, calculators, computer code, GIS mapping tools, practitioner guidelines, web applications, and various other toolkits and applied research products. All RAND tools undergo rigorous peer review to ensure both high data standards and appropriate methodology in keeping with RAND's commitment to quality and objectivity.
This document and trademark(s) contained herein are protected by law. This representation of RAND intellectual property is provided for noncommercial use only. Unauthorized posting of this publication online is prohibited; linking directly to this product page is encouraged. Permission is required from RAND to reproduce, or reuse in another form, any of its research documents for commercial purposes. For information on reprint and reuse permissions, please visit www.rand.org/pubs/permissions .
RAND is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.
IMAGES
COMMENTS
Examples of Specific Research Objectives: 1. "To examine the effects of rising temperatures on the yield of rice crops during the upcoming growth season.". 2. "To assess changes in rainfall patterns in major agricultural regions over the first decade of the twenty-first century (2000-2010).". 3.
Let's look at two examples of research objectives. We'll stick with the topic and research aims we mentioned previously. For the digital transformation topic: To observe the retail HR employees throughout the digital transformation. To assess employee perceptions of digital transformation in retail HR. To identify the barriers and ...
Formulating research objectives has the following five steps, which could help researchers develop a clear objective: 8. Identify the research problem. Review past studies on subjects similar to your problem statement, that is, studies that use similar methods, variables, etc.
Example: Research aim. To examine contributory factors to muscle retention in a group of elderly people. Example: Research objectives. To assess the relationship between sedentary habits and muscle atrophy among the participants. To determine the impact of dietary factors, particularly protein consumption, on the muscular health of the ...
Research Objectives. Research objectives refer to the specific goals or aims of a research study. They provide a clear and concise description of what the researcher hopes to achieve by conducting the research.The objectives are typically based on the research questions and hypotheses formulated at the beginning of the study and are used to guide the research process.
Both descriptive and analytical objectives are important in research as they serve different purposes and contribute to a comprehensive understanding of the research topic. Examples of Research Objectives. Here are some examples of research objectives in different fields: 1. Objective: To identify key characteristics and styles of Renaissance art.
Achievable: Objectives need to be realistic and attainable within the constraints of the research project, including time, resources, and expertise. Setting achievable goals prevents frustration and ensures steady progress. Relevant: Objectives must be aligned with the research problem and the overall purpose of the study.
Relevance: Your objectives should be relevant to your research topic and contribute to the existing knowledge in your field. Avoid objectives that are unrelated or insignificant, as they may waste your time or resources. ... Examples of Well-Written Research Objectives. Example 1- Psychology. Research question: What are the effects of social ...
A research aim is a broad statement indicating the general purpose of your research project. It should appear in your introduction at the end of your problem statement, before your research objectives. Research objectives are more specific than your research aim. They indicate the specific ways you'll address the overarching aim.
For example, the sample(s) to be approached for participation in the primary data collection. Identifying research objectives that are SMART is key to ensuring key aspects of the work are considered prior to any data collection. This includes consideration of access to samples and the ethics of researching the topic and research design.
5 Examples of Research Objectives. The following examples of research objectives based on several published studies on various topics demonstrate how the research objectives are written: This study aims to find out if there is a difference in quiz scores between students exposed to direct instruction and flipped classrooms (Webb and Doman, 2016).
Specificity: Objectives should be specific and narrowly focused on the aspects of the research topic that the study intends to investigate. They should answer the question of "what" or "which" rather than "how" or "why.". Measurability: Research objectives should be formulated in a way that allows for measurement and evaluation.
The first question asks for a ready-made solution, and is not focused or researchable. The second question is a clearer comparative question, but note that it may not be practically feasible. For a smaller research project or thesis, it could be narrowed down further to focus on the effectiveness of drunk driving laws in just one or two countries.
Summary. One of the most important aspects of a thesis, dissertation or research paper is the correct formulation of the aims and objectives. This is because your aims and objectives will establish the scope, depth and direction that your research will ultimately take. An effective set of aims and objectives will give your research focus and ...
To develop a set of research objectives, you would then break down the various steps involved in meeting said aim. For example: This study will investigate the link between dehydration and the incidence of urinary tract infections (UTIs) in intensive care patients in Australia. To achieve this, the study objectives w ill include:
A strong research topic comprises three important qualities: originality, value and feasibility.. Originality - a good topic explores an original area or takes a novel angle on an existing area of study.; Value - a strong research topic provides value and makes a contribution, either academically or practically.; Feasibility - a good research topic needs to be practical and manageable ...
Formulating research aim and objectives in an appropriate manner is one of the most important aspects of your thesis. This is because research aim and objectives determine the scope, depth and the overall direction of the research. Research question is the central question of the study that has to be answered on the basis of research findings.
Research Topic. Definition: Research topic is a specific subject or area of interest that a researcher wants to investigate or explore in-depth through research. It is the overarching theme or question that guides a research project and helps to focus the research activities towards a clear objective.
A well-crafted research question (or set of questions) sets the stage for a robust study and meaningful insights. But, if you're new to research, it's not always clear what exactly constitutes a good research question. In this post, we'll provide you with clear examples of quality research questions across various disciplines, so that you can approach your research project with confidence!
They outline what the researcher aims to achieve, helping to focus the research, design methodologies, and guide data collection and analysis. These objectives ensure the research stays on track and addresses key issues relevant to the topic. Examples of Research Objectives. To determine the impact of social media on adolescent mental health.
Research question. Interest in a particular topic usually begins the research process, but it is the familiarity with the subject that helps define an appropriate research question for a study. 1 Questions then arise out of a perceived knowledge deficit within a subject area or field of study. 2 Indeed, Haynes suggests that it is important to know "where the boundary between current ...
The choice between qualitative and quantitative research depends upon your research goals and objectives. Qualitative research methods are better suited when you want to understand the complexities of your user's problems and uncover the underlying motives beneath their thoughts, feelings, and behaviors.
A good research question is essential to guide your research paper, dissertation, or thesis. All research questions should be: Focused on a single problem or issue. Researchable using primary and/or secondary sources. Feasible to answer within the timeframe and practical constraints. Specific enough to answer thoroughly.
There is no single way to think of a research topic. Sometimes one just comes to you, but often you have to do some brainstorming and initial background research. Example. Imagine you are taking a psychology course, and your assignment is to write an essay exploring a topic related to phobias. You could start by looking in your course textbook ...
When they choose the topic of their final research projects, they take whatever data they gathered in a given week and present it in different formats, including film, poetry and op-eds.
In recent years, researchers have taken the many-objective optimization algorithm, which can optimize 5, 8, 10, 15, 20 objective functions simultaneously, as a new research topic. However, the ...
This codebook was compiled as part of the RAND Veterans Policy Research Institute's effort to develop a repository of datasets that assess veteran status. This codebook for the American Community Survey Public Use Microdata Sample (ACS-PUMS) presents an overview of the data, how to access them, which variable(s) assess veteran status, other ...
Here is an example of a resume objective vs a resume summary: ... Seeking a legal content writer role to utilize research and analytical skills in crafting compelling and legally sound content for digital platforms." "Marketing professional with a background in traditional advertising, transitioning to a career in digital marketing. ...