Modified Local Leader Phase Spider Monkey Optimization Algorithm

Abstract

An algorithm that modifies the local leader phase of the spider monkey optimization (SMO) algorithm is proposed. The proposed algorithm called modified local leader spider monkey optimization (MLLP-SMO) balances the search process in the local leader phase by offering chances to each spider monkey that is selected for update, to update to a better position, based on the strength of its previous fitness. The proposed algorithm was compared with the SMO and an improvement of the SMO called adaptive step-size based Spider Monkey Optimization (AsSMO), on nine benchmark problems. The comparison was done based on mean absolute error (MAE), standard deviation (SD) and convergence rate. The test results show that the MLLP-SMO performs better than the other two algorithms. The use of the proposed method in optimization problems will yield optimal values, with minimal iterations.