Genetic Programming for Cooperative Single-Objective Optimization
Keiko Ono, Masahito Kumano, Masahiro Kimura
We consider a method for solving cooperative single-objective optimization problems using genetic programming. The cooperative single-objective optimization problems are the special cases of single-objective optimization problems, where 1) multi-objective functions are defined, 2) the goal of this problem is to find optimal solution sets which each contain a different solution for each objective function, 3) cooperative evolving between objective-functions is effective for performance improvement. Many real-world systems involve cooperative single-objective optimization in their operations, so it is important to deal with this kind of optimization problems. In this paper, we first give the problem definition of a single-objective optimization problem, and then present a genetic programming method for a cooperative single-objective optimization problem which can collaborate among components in exchanging frequent tree structures extracted from good individuals.