An Artificial Decision Maker for Comparing Reference Point Based Interactive Evolutionary Multiobjective Optimization Methods
Afsar, B., Miettinen, K., & Ruiz, A. B. (2021). An Artificial Decision Maker for Comparing Reference Point Based Interactive Evolutionary Multiobjective Optimization Methods. In H. Ishibuchi, Q. Zhang, R. Cheng, K. Li, H. Li, H. Wang, & A. Zhou (Eds.), Evolutionary Multi-Criterion Optimization : 11th International Conference, EMO 2021, Shenzhen, China, March 28–31, 2021, Proceedings (pp. 619-631). Springer. Lecture notes in computer science, 12654. https://doi.org/10.1007/978-3-030-72062-9_49
Published inLecture notes in computer science
Li, Ke |
Li, Hui |
© Springer Nature Switzerland AG 2021
Comparing interactive evolutionary multiobjective optimization methods is controversial. The main difficulties come from features inherent to interactive solution processes involving real decision makers. The human can be replaced by an artificial decision maker (ADM) to evaluate methods quantitatively. We propose a new ADM to compare reference point based interactive evolutionary methods, where reference points are generated in different ways for the different phases of the solution process. In the learning phase, the ADM explores different parts of the objective space to gain insight about the problem and to identify a region of interest, which is studied more closely in the decision phase. We demonstrate the ADM by comparing interactive versions of RVEA and NSGA-III on benchmark problems with up to 9 objectives. The experiments show that our ADM is efficient and allows repetitive testing to compare interactive evolutionary methods in a meaningful way.
Parent publication ISBN978-3-030-72061-2
ConferenceInternational Conference on Evolutionary Multi-Criterion Optimization
Is part of publicationEvolutionary Multi-Criterion Optimization : 11th International Conference, EMO 2021, Shenzhen, China, March 28–31, 2021, Proceedings
ISSN Search the Publication Forum0302-9743
Publication in research information system
MetadataShow full item record
Related funder(s)Academy of Finland
Funding program(s)Academy Project, AoF; Research profiles, AoF
Showing items with similar title or keywords.
Desirable properties of performance indicators for assessing interactive evolutionary multiobjective optimization methods Aghaei Pour, Pouya; Bandaru, Sunith; Afsar, Bekir; Miettinen, Kaisa (ACM, 2022)Interactive methods support decision makers in finding the most preferred solution in multiobjective optimization problems. They iteratively incorporate the decision maker's preference information to find the best balance ...
Assessing the Performance of Interactive Multiobjective Optimization Methods : A Survey Afsar, Bekir; Miettinen, Kaisa; Ruiz, Francisco (Association for Computing Machinery (ACM), 2021)Interactive methods are useful decision-making tools for multiobjective optimization problems, because they allow a decision-maker to provide her/his preference information iteratively in a comfortable way at the same time ...
Towards explainable interactive multiobjective optimization : R-XIMO Misitano, Giovanni; Afsar, Bekir; Lárraga, Giomara; Miettinen, Kaisa (Springer Science and Business Media LLC, 2022)In interactive multiobjective optimization methods, the preferences of a decision maker are incorporated in a solution process to find solutions of interest for problems with multiple conflicting objectives. Since multiple ...
Comparing interactive evolutionary multiobjective optimization methods with an artificial decision maker Afsar, Bekir; Ruiz, Ana B.; Miettinen, Kaisa (Springer Science+Business Media, 2021)Solving multiobjective optimization problems with interactive methods enables a decision maker with domain expertise to direct the search for the most preferred trade-offs with preference information and learn about the ...
An Approach to the Automatic Comparison of Reference Point-Based Interactive Methods for Multiobjective Optimization Podkopaev, Dmitry; Miettinen, Kaisa; Ojalehto, Vesa (Institute of Electrical and Electronics Engineers (IEEE), 2021)Solving multiobjective optimization problems means finding the best balance among multiple conflicting objectives. This needs preference information from a decision maker who is a domain expert. In interactive methods, the ...