An alternative view for incorporating more scaled differences to differential evolution
Miguel Cárdenas‐Montes · Logic Journal of IGPL · 2018
Differential Evolution (DE) is one of the most popular metaheuristics. This is based on its performance and speed, and at the same time its easy description which allows direct modifications aimed at improving its performance. DE is based on the application of the scaled difference of a pair of population members to another population member, all of them distinct. Following this schema diverse variants have been proposed, progressively relaxing the constraints of the initial schema while improving the performance. In this work, new variants with a large number of scaled differences are proposed and evaluated against a large set of fitness functions, including shifted and rotated fitness functions. The proposed DE-variants demonstrate a high performance in the tests executed.