https://github.com/dibahk/uc-berkley-pacman-project

https://github.com/dibahk/uc-berkley-pacman-project

Last synced: about 1 month ago
JSON representation

• Host: GitHub
• URL: https://github.com/dibahk/uc-berkley-pacman-project
• Owner: dibahk
• Created: 2024-04-24T12:20:14.000Z (9 months ago)
• Default Branch: main
• Last Pushed: 2024-04-24T12:43:31.000Z (9 months ago)
• Last Synced: 2024-04-24T16:21:49.968Z (9 months ago)
• Language: Python
• Size: 2.27 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Pac-Man AI Projects

Welcome to the Pac-Man AI Projects repository! This collection of projects was developed for UC Berkeley's introductory artificial intelligence course, CS 188. As a former student who completed these projects, I can attest to their value in learning foundational AI concepts by applying various techniques to playing Pac-Man. Through these projects, I gained insights into informed state-space search, probabilistic inference, reinforcement learning, and more. These concepts are not just about gaming; they form the bedrock of real-world applications.

## Goals

Having completed these projects, I understand the three primary objectives they were crafted with:

1. **Visualization and Understanding**: The projects enabled me to visualize the outcomes of the techniques I implemented, fostering a deeper understanding of AI concepts.

2. **Clear and Concise**: I appreciated the clear directions and code examples provided, which helped me grasp the concepts without being overwhelmed by excessive scaffolding.

3. **Challenge and Creativity**: Pac-Man provided a challenging problem environment that encouraged me to devise creative solutions, mirroring the complexity of real-world AI problems.

As someone who has completed these projects, here's a brief overview:

**Search**: Implement depth-first, breadth-first, uniform cost, and A* search algorithms to solve navigation and traveling salesman problems in the Pac-Man world.

**Multi-Agent Search**: Model classic Pac-Man as both an adversarial and stochastic search problem. Implement multi-agent minimax and expectimax algorithms, along with designing evaluation functions.

**Reinforcement Learning**: Implement model-based and model-free reinforcement learning algorithms, applied to various environments including Gridworld, Pac-Man, and a simulated crawling robot.

**Ghostbusters**: Explore probabilistic inference in a hidden Markov model to track the movement of hidden ghosts in the Pac-Man world. Implement exact inference using the forward algorithm and approximate inference via particle filters.

Having walked the path through these projects, I encourage you to dive in and explore the exciting world of AI through Pac-Man! If you have any questions or encounter issues, don't hesitate to reach out. Happy coding!