Yadav, Himanshu2025-02-052025-02-052024-0656p.http://hdl.handle.net/10263/7502Dissertation under the supervision of Dr. Swagatam DasRecent advancements in Monte Carlo Tree Search (MCTS) methods have garnered significant attention in the AI community. Inspired by these developments, we implemented tree-based search algorithms to create game-playing agents for Tic-Tac-Toe and Ludo. Our analysis primarily focuses on the Minmax and MCTS algorithms. Minmax, a well-known strategy in game theory, effectively handles games with smaller state spaces like Tic-Tac-Toe, ensuring optimal play. However, for more complex games like Ludo, MCTS proved to be more efficient by using random sampling and tree-based search to make decisions. We further explored Bayesian MCTS for Tic-Tac-Toe, which incorporates probabilistic models to enhance decision-making. This approach allowed our agent to handle uncertainty better, leading to improved performance. These implementations demonstrate the potential of combining traditional AI techniques with advanced search algorithms to create robust game-playing agents.enTic-Tac-ToeLudo GameGame theoretic propertiesAI Game Playing agentsBayesian InferenceOther