QPSO algorithm based on Lévy flight and its application in fuzzy portfolio

Abstract

To solve constrained portfolio selection model effectively, an improved quantum-behaved particle swarm optimization algorithm(LQPSO) is presented. Firstly, considering its practicality in real dealing process, a class of fuzzy portfolio models with transaction costs and background risk is established. Then in the design of improved algorithm, Lévy flight strategy and contraction–expansion coefficient with nonlinear structure are taken into account for enhancing particle’s exploration ability, and premature prevention mechanism is used to increase population diversity. According to the following performance test, LQPSO demonstrates better convergence and robustness than PSO with inertia weight, QPSO and QPSO with a hybrid probability distribution in 12 benchmark functions. Furthermore, experimental results indicate that LQPSO outperforms several metaheuristics when seeking optimal solution for the fuzzy portfolio model with constraints.