Index

In this project, I followed the instruction and gradually implemented the required function in ray tracing. The main programming language is C++, and OPENGL and GLUT are utilized to simulate the scene and object.

1. Search in Pacman
2. Multi-Agent Pacman
3. Reinforcement Learning
4. Ghostbusters

Search in Pacman

For pacman, find the efficient way to collect food in the maze without ghost.

• Python
• Method: DFS, BFS, A*, and UCS.
• Design effective heuristics function.

Multi-Agent Pacman

Allow pacman to collect all food without being eaten by ghost.

• Python
• Decision making strategy: Minimax, Alpha-Beta Pruning, and Expectimax.

Reinforcement Learning

Train pacman by real interaction experience, and in the end test the stability of the learning result.

• Python
• Method: Markov decision processes, Q-Learning, value iteration, and Approximate Q-Learning.

Ghostbusters

Design Pacman agents that use sensors to locate and eat invisible ghosts.

• Python
• Method: Particle Filter