Snake Game AI: A Comparative Study of A-Star Algorithm and Deep-Q Neural Network Approaches

About Our Study

This codebase is part of an in-depth research study examining deterministic and non-deterministic approaches for training automated agents in playing the Snake game. Specifically, our study compares the performance of the A* algorithm with a deep-q neural network.

Detailed Research Findings

For detailed insights, analysis, and comprehensive results, explore the full research paper available in the "Results" folder of this repository. The research paper provides extensive details on our experimental methodologies, findings, and conclusions drawn from the comparison between the A* algorithm and the deep-q neural network in playing the Snake game.

Running the Code

To run the code, you will need to install the following dependencies as specified in the requirements.txt file:

The main dependencies are:

• Python (3.11.5)
• Pygame
• PyTorch
• MatPlotLib.pyplot
• IPython

To run the Deep-Q Neural Network you need to run the agent.py file in the Deep-Q directory. To run the A-Star Algorithm you need to run the game.py file in the A-Star directory. If you want to play the game yourself without any AI, you can run the snake_game_human.py file in the root directory.

The Snake Game Implementation Credit

The Snake game implementation found here is originally credited to Patrick Loeber. You can explore the original implementation here or watch a detailed explanation in this video.