A Novel PSO-Inspired Probability-based Binary Optimization Algorithm

Abstract

Particle swarm optimization (PSO) is an emerging intelligent optimization algorithm. Because of its excellent characteristic, PSO has been wildly researched and applied to tackle optimization algorithm in continuous space. As PSO cannot optimize the discrete optimization problem, Kennedy proposed a discrete binary PSO firstly, and then Shen proposed a modified binary PSO, but the further research and application works are few as the optimization ability of binary PSO is not ideal. To tackle binary optimization problems more effectively, we propose a novel probability binary PSO algorithm based on PSO and probability optimization algorithm, called PSO-inspired probability-based binary optimization algorithm (PPBO). The experimental results demonstrate that the proposed PPBO is valid and outperforms the discrete binary PSO and the modified binary PSO in terms of the optimization efficiency and ability.