摘要：This paper deals with fuzzy multi-objectivemulti-period portfolio selection problems. The portfolioselection is proposed by taking into account three criteria offinal return, cumulative risk and entropy. In the model, thereturn level is quantified by the possibilistic mean value ofreturn, and the risk is quantified by the possibilistic variance ofreturn while fuzzy entropy is adopted to increase the riskdispersion degree to some extent. Then a fuzzy multi-objectivemulti-period portfolio model is presented in a more complexmarket environment. To solve the complex model, themulti-objective functions are transformed into a single objectiveand the risk preference parameter is introduced to balance thereturn and risk to meet with investors’ preferences. To ensurethe investor can obtain the optimal portfolio strategy, a hybridintelligent algorithm is designed by combining both geneticalgorithm and wavelet neural network algorithm, which notonly utilizes the good localization property of wavelet transformbut also utilizes the effective self-learning function of neuralnetwork. Finally, a numerical example is presented to illustratethis approach and the designed algorithm. The results show thatthe proposed model and the designed algorithm are practicaland flexible, while they are meaningful for the study onportfolio selection and multi-objective programming.
