Comparison of Multi-agent Co-operative Co-evolutionary and Evolutionary Algorithms for Multi-objective Portfolio Optimization

Co-evolutionary techniques makes it possible to apply evolutionary algorithms in the cases when it is not possible to formulate explicit fitness function. In the case of social and economic simulations such techniques provide us tools for modeling interactions between social or economic agents--especially when agent-based models of co-evolution are used. In this paper agent-based versions of multi-objective co-operative co-evolutionary algorithms are applied to portfolio optimization problem. The agent-based algorithms are compared with classical versions of SPEA2 and NSGA2 multi-objective evolutionary algorithms.