Group search optimizer algorithm with predictive model
Zhaohui Zhu · Information technology newsletter · 2010
An improved GSO(Group Search Optimizer) is presented in this paper.The major improvement is that swarm members save good direction of movements during the search process as their experience and predict better positions from these experiences.The improved GSO has a superior search performance on low-dimensional problems and high-dimensional problems.The improved GSO algorithm is both simpler and faster than GSO algorithm.A set of 4 benchmark functions were employed to evaluate the improved algorithm.For the 30 dimensional cases,the improved algorithm outperformed GA and GSO for all the 4 benchmark functions,and outperformed PSO for 2 of them.For the 300 dimensional cases,the improved algorithm has a markedly superior performance to GA,PSO and GSO algorithms.