Keywords
How to Cite
Abstract
This research aims to evaluate the optimum portfolio selection using with particle swarm algorithm and genetic algorithm. For this purpose, the financial information of companies listed on the Iran stock exchange, during years 2007 to 2012 is collected and using heuristic particle swarm algorithm, genetic algorithms and based on Markowitz model, mean-variance model and client risk model, generating optimal portfolio from the stocks has been investigated. In total, the results of this study showed that use of these algorithms can provide solutions both close together and close to optimality, and causes confidence of the investors' investment for making decisions. Also, based on the response obtained by performing several experiments it can be claimed that in Markowitz and mean-variance models can provide most optilam portfolio. In other hands, particle swarm algorithm is best in client risk model. Most observations reflect the fact that in the problems which are smaller and lighter the genetic algorithm, and as the complexity and size increases, the particle swarm algorithm perform better.