https://github.com/sameerjain/ai-snake-game
Implemented a Q-learning agent to play and learn the classic Snake game, leveraging PyTorch for Neural Network. Visualized using Matplotlib, showcasing agent's evolving performance over time. Built UI using Pygame to visualize Snake, including food, and current score.
https://github.com/sameerjain/ai-snake-game
anaconda conda matplotlib neural-network numpy pygame python3 pytorch q-learning
Last synced: 5 months ago
JSON representation
Implemented a Q-learning agent to play and learn the classic Snake game, leveraging PyTorch for Neural Network. Visualized using Matplotlib, showcasing agent's evolving performance over time. Built UI using Pygame to visualize Snake, including food, and current score.
- Host: GitHub
- URL: https://github.com/sameerjain/ai-snake-game
- Owner: SameerJain
- Created: 2023-11-13T17:10:44.000Z (about 2 years ago)
- Default Branch: main
- Last Pushed: 2024-09-12T04:04:20.000Z (over 1 year ago)
- Last Synced: 2025-04-02T18:18:03.598Z (10 months ago)
- Topics: anaconda, conda, matplotlib, neural-network, numpy, pygame, python3, pytorch, q-learning
- Language: Python
- Homepage: https://www.youtube.com/watch?v=RaI2uuSOi6c
- Size: 79.1 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
**Video Demonstration:**
https://www.youtube.com/watch?v=RaI2uuSOi6c
# Snake Game Q-Learning AI
This project implements an AI agent that learns to play the classic Snake game using Q-learning, a type of reinforcement learning algorithm. The AI improves its performance over time by learning from its experiences in the game. NOTE: This uses Anaconda
## Requirements
Python 3.7+
Anaconda
PyTorch
Pygame
Matplotlib
NumPy
## How to get started
1) Download the zip and extract the file
2) Run Anaconda:
`conda create -n pygame_env`
3) Run the AI Agent:
`python agent.py`
4) Watch the AI learn!
## Project Structure
**snake_game.py:** Implements the Snake game using Pygame.
**model.py:** Defines the neural network model for Q-learning.
**agent.py:** Implements the Q-learning agent that interacts with the game.
**plot.py:** Provides functionality to visualize the training progress.
### Contains:
Game initialization and reset
Food placement
Game state updates
Collision detection
UI rendering
## How it Works
The Q-learning process works as follows:
**State Representation:**
The game state is represented as a vector of 11 binary values in get_state() method.
This includes information about danger directions, current direction, and food location.
**Q-Network:**
The Q-function is approximated using a neural network (Linear_QNet class).
It takes the state as input and outputs Q-values for each possible action.
**Action Selection:**
The agent uses an epsilon-greedy strategy in get_action() method.
With probability epsilon, it chooses a random action (exploration).
Otherwise, it chooses the action with the highest Q-value (exploitation).
Epsilon decreases over time to favor exploitation as the agent learns.
**Learning Process:**
After each action, the agent receives a reward and observes the new state.
The experience (state, action, reward, next_state) is stored in memory.
The Q-network is updated using experience replay in train_long_memory():
A batch of experiences is randomly sampled from memory.
The Q-values are updated using the Bellman equation:
Q(s,a) = r + γ * max(Q(s',a'))
The network is trained to minimize the difference between predicted and target Q-values.
**Game Integration:**
The SnakeGameAI class provides the environment for the agent.
In each step, the game updates based on the agent's action and returns the new state and reward.
**Visualization:**
The plot() function provides real-time visualization of the agent's performance.
It shows how the score improves over time as the agent learns.