Skip to content

A team of capture agents for the Pacman Capture the Flag challenge using Monte Carlo Tree Search. The project implements offensive and defensive strategies with MCTS to optimize real-time decision-making in game scenarios, building on the UC Berkeley Pacman AI framework.

License

Notifications You must be signed in to change notification settings

Evan09064/MCTS-Based-Capture-Agents-for-Pacman-AI

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

MCTS-Based Capture Agents for Pacman AI

This project implements a team of capture agents for the Pacman Capture the Flag challenge using Monte Carlo Tree Search (MCTS). The agents are divided into two roles:

  • OffensiveMCTSAgent: Uses MCTS to plan aggressive moves when pursuing food and evading enemies.
  • DefensiveMCTSAgent: Uses MCTS and patrol strategies to defend territory and intercept invaders.

Features

  • MCTS Implementation: Includes node expansion, simulation, evaluation, and backpropagation to guide decision-making.
  • Dual Role Agents: Separate offensive and defensive strategies tailored to in-game situations.
  • Seamless Integration: Designed to work within the UC Berkeley Pacman AI projects framework.

Setup

  1. Clone the Repository:

  2. Integration:

  • Place the source file (e.g., GEAK (3).py) in your Pacman project’s directory where team agents are defined.
  • Ensure the project includes the original UC Berkeley files (e.g., captureAgents.py, game.py, etc.) to support integration.
  1. Usage:
  • The team is created using the createTeam function defined in the file.
  • Run the Pacman Capture the Flag game and select this team configuration to see the agents in action.

Attribution

This project is based on the UC Berkeley Pacman AI projects. Please retain the original licensing and attribution information as provided.

License

[MIT License]

About

A team of capture agents for the Pacman Capture the Flag challenge using Monte Carlo Tree Search. The project implements offensive and defensive strategies with MCTS to optimize real-time decision-making in game scenarios, building on the UC Berkeley Pacman AI framework.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages