Information guiding and sharing enhanced simultaneous heat transfer search and its application to k-means optimization

Abstract

Simultaneous Heat Transfer Search (SHTS) is a novel meta-heuristic algorithm proposed recently and it can solve some optimization problems. However, it still has some deficiencies, such as weak search ability owing to poor information exploitation so that it cannot solve complicated problems well. So an Information Guiding and Sharing enhanced SHTS (IGS-SHTS) is proposed in this paper. Firstly, Grey wolf optimizer is embedded into SHTS to enhance information sharing through the two different search methods. Secondly, an information sharing way through different agents is added. Thirdly, a sinusoidal crossover is utilized to fulfill information guiding through the historical individual-best population guiding the current one. Finally, an information guiding way, the global best agent guiding the worst one, is embedded to strengthen the worst agent. A large number of experimental results on the complex functions from CEC2017 test set and clustering optimization show that IGS-SHTS has stronger search ability compared with SHTS and quite a few state-of-the-art algorithms.