Improved Chaotic Grey Wolf Optimization for Training Neural Networks

Abstract

This paper introduces one improved version of the Grey Wolf Optimization algorithm (GWO), one of the newly established nature-inspired metaheuristic algorithms, and the suggested approach is termed Chaotic Grey Wolf Optimization (CGWO). The newly suggested approach CGWO is premeditated by the integration of the chaos technique with the GWO algorithm, aiming to resolve global optimization problems by maintaining a proper balance between exploration and exploitation. In the proposed approach, CGWO is assessed over the classic 23 benchmark functions. The proficiency of the freshly suggested approach, CGWO is verified by comparing it with contemporary methods as well as examined through statistical analysis also. Further, the same CGWO is utilized to train neural networks (MLP) by considering benchmark datasets, for data classification and establishing a better classifier algorithm.