Improved particle swarm algorithm for portfolio optimization problem

Particle swarm optimization (PSO) is a recently proposed population-based random search algorithm, which performs well in some optimization problems. In this paper, we proposed an improved PSO algorithm to solve portfolio selection problems. The proposed approach IPSO employs an opposite mutation operator to enhance the performance of the standard PSO. In order to verify the performance of IPSO, we test it on five well-known benchmark function optimization problems. At last, we use IPSO to solve a classical portfolio selection problem. The results show that the proposed approach is effective and achieves better results than standard PSO.