A project that delves deep into game theoretical models and probabilistic decision-making through the classic game of Pacman.
This project focuses on implementing and analyzing various AI agents that play Pacman. By applying concepts from game theory and probabilistic models, we can observe how these agents react to different scenarios and adversaries.
Reflex Agent: An agent that considers both food locations and ghost locations to make decisions.
Minimax Algorithm: Adversarial search algorithm for decision-making in multi-agent settings like Pacman.
Alpha-Beta Pruning: Optimization on Minimax, which prunes the search tree to explore the most promising moves.
Expectimax Algorithm: Accounts for probabilistic behavior of agents, providing an expectation of all possible outcomes.