On the performance of Recurring Multistage Evolutionary Algorithm for continuous function optimization

Recurring Multistage Evolutionary Algorithm is a novel evolutionary approach that is based on repeating conventional, explorative and exploitative genetic operations in order to perform better optimization with improved robustness against local optima. This work compares the performance of RMEA with that of classical evolutionary algorithm, differential evolution and particle swarm optimization on a test suite of 50 different benchmark functions. The test functions include unimodal and multimodal, separable and non-separable, regular and irregular, low and high dimensional functions. Very few works have been tested on a similar range of benchmark problems. The experimental results show that the performance of RMEA is comparable to and often better than the other mentioned algorithms.