Structure-Redesign-Based Bacterial Foraging Optimization for Portfolio Selection

摘要

In this paper structure-redesign-based Bacterial Foraging Optimization (SRBFO) is proposed to solve portfolio selection problem. Taking advantage of single-loop structure, a new execution structure is developed in SRBFO to improve the convergence rate as well as lower computational complexity. In addition, the operations of reproduction and elimination-dispersal are redesigned to further simplify the original BFO algorithm structure. The proposed SRBFO is applied to solve portfolio selection problems with transaction fee and no short sales. Four cases with different risk aversion factors are considered in the experimental study. The optimal portfolio selection obtained by SRBFO is compared with PSOs, which demonstrated that the validity and efficiency of our proposed SRBFO in selecting optimal portfolios.