A Novel Multi-objective Evolutionary Algorithm Solving Portfolio Problem

Abstract

With the improvement of complex and uncertain finance environment,the difficulty of portfolio problem is increasing. Whether or not the projects is successfully selected, directly affects the development of the investment companies. This paper firstly talks about the finance conditions in single term investment and then extends the investment from one term to many terms. After that, a multi-project and multi-term portfolio model through considering the remaining funds in different investment terms is proposed. The model is based on a new kind of Mean-Semi-covariance theory, which can describe the uncertainty of return and risk in investment. The portfolio investment is a multi-objective optimization problem with constraints. Multi-objective evolutionary algorithm (MOEA) with greedy repair strategy is used to deal with the infeasible individuals and makes the investment reasonable. Finally, computer simulation shows that the proposed algorithm can be considered as a viable alternative.