Interactive Approaches Applied to Multiobjective Evolutionary Algorithms
Antonio López Jáimes, Carlos A. Coello Coello · 2013
Multiobjective optimization involves three stages: model building, search, and decision making (preference articulation). This chapter presents a review of recent MOEAs designed to work as interactive optimization methods. This chapter aims to complement previous reviews of the field instead of aiming at being comprehensive. By analyzing the approaches found in other reviews and the ones covered in this chapter, one can realize that most of the approaches to incorporate preferences into MOEAs are based on methods introduced in the field of multicriteria decision making. The methods analyzed in this chapter are classified into the following categories: reference point methods, utility function methods and miscellaneous methods. The survey is mainly focused on interactive MOEAs. The chapter also includes some of these interesting a priori techniques for incorporating preferences. Controlled Vocabulary Terms decision making; operations research