Pac-Man AI
Intelligent AI agents designed to master the classic Pac-Man game using advanced algorithms including search, adversarial agents, and reinforcement learning.
📚 Academic Project - Repository and demo will be available soon
AI Techniques Implemented
Academic Excellence
This project was developed as part of advanced coursework in Artificial Intelligence. My implementation achieved exceptional performance in the competitive tournament.
Technologies & Algorithms
Designed intelligent AI agents to master the classic Pac-Man game through multiple AI paradigms. This comprehensive project implements various artificial intelligence techniques to create sophisticated agents capable of strategic gameplay, learning, and adaptation.
Search Algorithms
- •A* algorithm for optimal pathfinding
- •Custom heuristics for maze navigation
- •Real-time search optimization
- •DFS & BFS implementation
Adversarial Agent
- •Alpha-beta pruning for competition
- •Minimax strategic decision making
- •Multi-agent ghost scenarios
- •Game tree exploration
Reinforcement Learning
- •Q-learning strategy adaptation
- •Value Iteration for optimal policy
- •Reward-based navigation
- •Temporal difference learning
Perceptron Model
- •Custom perceptron from scratch
- •Move prediction implementation
- •Feature extraction & weights
- •Real-time decision making
Academic Project: This project demonstrates mastery of fundamental AI concepts through practical implementation in a classic gaming environment.