Zastosowanie koewolucyjnego algorytmu genetycznego w rozwiązaniu zadania trójkryterialnego
Application of co-evolutionary genetic algorithm for a three-criterion problem

Summary/Abstract: In this work we present a relatively simple method of adapting a genetic algorithm to a triple-criterion problem by using co-evolution. By running the calculations for multiple − distinctly different yet connected, sub-populations we can easily generate a set of Pareto-optimal solution for a multi-criterion problem. The proposed method is tested for a portfolio optimization algorithm using expected returns as well as upper and lower semivariance. The results prove that the proposed method is an effective tool for multi-criteria optimisation of a numerically diffiucult functions.

• Details
• Contents