Differential evolution for high-dimensional function optimization

Zhenyu Yang, Ke Tang, Xin Yao

Abstract: Most reported studies on differential evolution (DE) are obtained using low-dimensional problems, e.g., smaller than 100, which are relatively small for many real-world problems. In this paper we propose two new efficient DE variants, named DECC-I and DECC-II, for high-dimensional optimization (up to 1000 dimensions). The two algorithms are based on a cooperative coevolution framework incorporated with several novel strategies. The new strategies are mainly focus on problem decomposition and subcomponents cooperation. Experimental results have shown that these algorithms have superior performance on a set of widely used benchmark functions.