A competitive learning-based Grey wolf Optimizer for engineering problems and its application to multi-layer perceptron training

Abstract

This article presents a competitive learning-based Grey Wolf Optimizer (Clb-GWO) formulated through the introduction of competitive learning strategies to achieve a better trade-off between exploration and exploitation while promoting population diversity through the design of difference vectors. The proposed method integrates population sub-division into majority groups and minority groups with a dual search system arranged in a selective complementary manner. The proposed Clb-GWO is tested and validated through the recent CEC2020 and CEC2019 benchmarking suites followed by the optimal training of multi-layer perceptron’s (MLPs) with five classification datasets and three function approximation datasets. Clb-GWO is compared against the standard version of GWO, five of its latest variants and two modern meta-heuristics. The benchmarking results and the MLP training results demonstrate the robustness of Clb-GWO. The proposed method performed competitively compared to all its competitors with statistically significant performance for the benchmarking tests. The performance of Clb-GWO the classification datasets and the function approximation datasets was excellent with lower error rates and least standard deviation rates.

MLBRSA: Multi-Learning-Based Reptile Search Algorithm for Global Optimization and Software Requirement Prioritization Problems

“…Therefore, it is essential to note that while the RSA has these potential defects, it also has strengths, and its performance can be problem-dependent. The proposed enhancements, including the integration of Q-learning, competitive learning, and adaptive learning, aim to address some defects and improve the algorithm's robustness and efficiency [81,82].…”

“…In this study, competitive learning influences how solutions are updated. The winning solution (the one with the best fitness) provides a reference or guide for updating other solutions [81]. This ensures that (i) the search is biased towards regions of the search space that have yielded good solutions, (ii) solutions can escape local optima by being influenced by the global best or other high-performing solutions, and (iii) the diversity of solutions is maintained, as not all solutions are pulled towards the best one, but are updated with a mix of exploration and exploitation.…”