Development of an AI-Powered Chess Engine Using Minimax Algorithm and Genetic Algorithm for Evaluation Function

Abstract- This project demonstrates a high level processing chess engine employing the Minimax algorithm along alpha-beta pruning, one more added feature used is a genetic algorithm which proves useful to make decisions and performance higher. While the Minimax algorithm is a cornerstone of game theory, which helps one to discover best moves and counter-moves in order not to lose in games like chess, with Alpha-beta pruning you can limit the number of nodes that are evaluated and hence restrict computational power needed without loosing optimality. Our evaluation function rates board states, based on which we use a genetic algorithm to fine-tune it. The optimal criteria are formed by the selection and combination of those evaluation functions over generations, while the genetic algorithm evolves a population of candidate solutions. This continuous refinement allows the evaluation function to improve as it gives a better result. While playing, the engine uses the so-called Minimax algorithm with alpha-beta pruning to look ahead and move sequences up to a certain depth for better decision-making. We tackle both tactical and strategic parts of chess in our implementation, showing strong play against humans. The project has had an analysis, which shows that the move selection and game outcomes are superior to conventional Minimax-based engines. This breakthrough in the class of Minimax algorithms achieves higher intelligence levels in computer chess, drastically changing gameplay for both fun and competitive purposes.