Combing Gibbs-sampling with Adaptive Particle Swarm for Large Scale Global Optimization

A Large Scale Global Optimization (LSGO) problem means the problem has over hundreds of decision variables. Many of the problems in the real world have such attributes, so how to effectively solve the LSGO problem has attracted the attention of many researchers. One of the biggest difficulties for the LSGO lies in the exponential growth of search space as variables increase. In this paper, we combine Gibbs Sampling with Adaptive Particle Swarm optimization algorithm (APS), and then propose a novel LSGO approach called Gibbs-APS. Our basic idea is using a multivariate Gaussian distribution to model the distribution of the particle swarm, such that to obtain better LSGO solutions. We compare the proposed method with four large scale global optimization algorithms on fifteen different test instances. The experimental results affirm the effectiveness of the proposed method in addressing large scale global optimization problems.