An Analysis on the Effectiveness of Adaptive MCST-CE on Optimizing Neural Network Weights

In this study, we explore the potential of integrating a search strategy with chaotic evolution to navigate the complete genetic tree structure.Our focus in this paper is on optimizing neural network weight.Chaotic evolution(CE) creates a tree of possible states and utilizes a fitness function to provide rewards.Given that Monte Carlo Tree Search (MCTS) has demonstrated effectiveness in managing states and rewards, we aim to combine these methodologies to effectively search for the best outcomes generated by chaotic evolution algorithms.From the prior research[11], we saw that the fusion of chaotic systems really made sense in enhancing the performance of the traditional CE.By combining these algorithm with the niching method and the MCTS, we will examine if the proposed algorithm can get the most optimal neural network weights.