🕹️👻👾👻 In this thrilling AI adventure, we embark on a multi-stage quest to transform Pacman into an intelligent game-playing agent. Across three engaging projects, we explore various facets of artificial intelligence, from basic search algorithms to adversarial competition and reinforcement learning.
Our journey begins with Pacman navigating complex mazes. We kick things off by implementing fundamental search algorithms like Depth-First Search and Breadth-First Search. As Pacman quests for optimal paths, we study the intricacies of uninformed search. Then, we take it up a notch with informed search algorithms, employing A* Search with different heuristics to guide Pacman through challenging terrains.
In the next phase, Pacman faces off against adversarial agents in a high-stakes game. We delve into adversarial search algorithms, unleashing the power of Minimax and Alpha-Beta Pruning. As Pacman engages in strategic warfare, we analyze the performance of diverse evaluation functions, helping him make calculated moves and outsmart his opponents.
In the final leg of our journey, we train Pacman to evolve into a masterful player through reinforcement learning. We implement cutting-edge algorithms like Q-Learning and Approximate Q-Learning, enabling Pacman to learn from experience and make intelligent decisions. Along the way, we explore the impact of feature design and parameter tuning on Pacman's learning, transforming him into an AI powerhouse.