Benchmarking SHADE algorithm enhanced with model based optimization on the BBOB noiseless testbed
Michał Okulewicz, Mateusz Zaborski · Proceedings of the Genetic and Evolutionary Computation Conference Companion · 2021
In this paper we evaluate the SHADE-LM algorithm on the BBOB noiseless testbed. The algorithm hybridizes the SHADE algorithm with a model based optimization. This hybridization is performed in a transparent manner for both optimizers, with SHADE having access to the samples provided by model based optimization, and models of square functions are fitted on the current population. The paper compares this extended version with the performance of the version of SHADE by Tanabe and Fukunaga.