Fuzzy Portfolio Optimization using Meta-heuristic Unconscious Search Algorithm

The optimization of the stock portfolio and the allocation of wealth between the various assets are among the most important issues in investing.In this study, the problem of optimizing stock portfolios, considering the real world constraints and with the assumption that the return on risky assets of fuzzy numbers is composed. Then, a new probabilistic model of mean-semi absolute deviation was presented in which transaction cost and cardinality constraints were also considered. The existence of such constraints transforms the model into a mixed-integer non-linear programming model that traditional approaches fail to solve, for this purpose a new meta-heuristic algorithm called the Unconscious Search Algorithm is used to solve the problem. Unconscious Search algorithm is a new algorithm based on Freud's psychoanalysis theory. In order to investigate the power and accuracy of solving this algorithm, a case study was carried out with the information of 50 top Tehran Stock Exchanges for years 2012 to 2016. The results were compared with Particle Swarm Optimization and Genetic algorithms, which showed the superiority of this algorithm in the optimization problem of Stock portfolio.