Portfolio analysis based on multi-objective optimization algorithm

Evolutionary algorithm in risk minimization and the expected return maximization of bi-objective for portfolio optimization applications has received widespread attention. Although the problem is a quadratic programming (QP) problem, the practical investment problems tend to make variables discontinuous and introduce other complexities. Under this circumstance, a normal QP solution is not always capable of finding a feasible solution. In this paper the NSGA-II algorithm is used to deal with the situation for classical QP unconventional methods. Results demonstrate that the evolutionary algorithm NSGA-II can find out the front of the conflict optimization problem which is difficult to achieve through other methods.