A new evolutionary approach using pre-post testing to trigger exploration and exploitation in DOPs
Hajer Ben-Romdhane, Enrique Alba, Saoussen Krichen · Proceedings of the Genetic and Evolutionary Computation Conference Companion · 2017
Striking an effective balance between exploration and exploitation (E&E) is still one of the major concerns when using evolutionary algorithms (EAs) in dynamic environments. In this work, a new scheme for adaptively balancing E&E in EAs is proposed. Based on the results of a statistical Pre-Post analysis of the population, the next search mode can be decided (i.e., exploration or exploitation). The experimental results showed that our proposal excels versus several competing approaches from the state of the art.