Global optimization via differential evolution with automatic termination
Bun Theang Ong, Masao Fukushima · Numerical Algebra Control and Optimization · 2012
Evolutionary Algorithms (EAs) provide a very powerful tool forsolving optimization problems. In the last decades, numerousstudies have been focusing on improving the performance of EAs.However, there is a lack of studies that tackle the question ofthe termination criteria. Indeed, EAs still need terminationcriteria prespecified by the user. In this paper, we propose tocombine the Differential Evolution (DE) method with novelelements, i.e., the ``Gene Matrix'' (GM), the ``SpaceDecomposition'' (SD) and ``Space Rotation'' (SR) mechanisms, inorder to equip DE with an automatic termination criterion withoutresort to predefined conditions. We name this algorithm``Differential Evolution with Automatic Termination'' (DEAT).Numerical experiments using a test bed of widely used benchmarkfunctions in 10, 50 and 100 dimensions show the effectiveness ofthe proposed method.