USING MACHINE LEARNING TO MODEL THE BEHAVIOR OF AN INTELLIGENT AGENT IN A STRATEGIC GAME

The creation of intelligent agents for board games is based on various decision-making methods. The article discusses the minimax algorithm and the Monte Carlo method, which are used in well-known intellectual games such as chess and go. The article includes a description of machine learning algorithms that make it possible to obtain an agent capable of making profitable decisions for himself in a short time, based on the condition assessment function. The paper presents the results of testing the obtained intelligent agents.