A Review towards human intuition based chess playing system using AI & ML

Abstract

This paper reviews the tools and strategies utilized for plan and advancement of chess systems directly from the time of Alan Turing to the new improvements in AlphaZero framework. Many best in class chess systems depended upon more conventional search methods like Alpha beta and Minimax till the new accomplishment of frameworks which depend on Artificial intelligence and machine learning idea. Advancements of the framework dependent on conventional search methods is conceivable simply by expansion of more processors to a biggest cutoff, accelerating action of fundamental modules using faster multiprocessors and upgrading the level of chess program especially its heuristic assessment function. Many chess systems are developed so as to not play instinctively like people, but figure out much varieties as could be expected and base their choice upon that. Latest improvements in game systems figured out how AlphaZero utilized new methodology dependent on reinforcement learning. It was able to learn without heuristic information about the game by playing a wide number of games with itself and was able to outplay traditional chess engines. Chess engines can be further improved by using hybrid evolutionary algorithms.