Improving global optimization ability of GSO using ensemble learning

As a novel bionic swarm intelligence optimization method, Glowworm Swarm Optimization (GSO) algorithm is inspired by the social behavior of glowworm and the phenomenon of bioluminescent communication, but GSO is easy to fall into local optimization point, and has the low speed of convergence in the late. In order to solve these problems, a method GSOE, combined with the GSO and the ensemble learning method, is presented. Through 4 typical functions testing, experiment results show that the method offers an effective way to avoid local optimization, and can improve the optimization global ability obviously.