Overview

In this coursework, you'll be implementing a Q-learning algorithm to control Pacman in the classic game environment. The aim is to enable Pacman to make intelligent decisions on how to move through the game. The effectiveness of your Q-learning algorithm will be assessed based on Pacman's performance in playing the game.

Task Description

Part 1: Getting Started

• Download the provided Pacman code from the specified link.
• Set up the environment and familiarize yourself with running Pacman using Python 3.
• Understand the command-line options and how to execute the game with different configurations.

Part 2: Towards an RL Pacman

• Implement the Q-learning algorithm in the provided mlLearningAgents.py file.
• The QLearnAgent class includes methods such as init(), getAction(), and final(), which you'll need to modify to incorporate Q-learning functionality.
• Ensure that your implementation allows Pacman to learn and make decisions based on learned utilities, balancing exploration and exploitation.