Huligutta

Project advisor: Narayana Prasad Santhanam, Associate Professor

Motivation

Develop a Reinforcement Learning algorithm on strategy games.

To run,

python game.py

Project Structure

├── README.md
├── dataset                 # contains generated dataset
│   └── data.txt
├── functions.py
├── game.py                 # Handles the GUI
├── huligutta.py            # Game code
├── images
├── notebooks
│   ├── Playground.ipynb    # Experimental notebook
│   └── RL.ipynb            # Some data visualizations
└── references

How to Play

Goats

• Click any empty positions to place a goat on the board
• To move goats, click the goat, then click on a valid empty position

Tigers

• To move tigers, click the tiger, then click on a valid empty position
• To capture, click on a valid empty position the tiger goes to.

Todo

• Play the game so it generates data.
• Computer vs computer
• Develop Reinforcement learning scheme
• Undo move feature.
• Organize how the data.txt collects data (i.e. what other valuable information are needed).

Dependencies

• networkx==2.5
• pillow==8.0.1
• numpy==1.19.2
• scipy==1.5.2

References

how-to-play-goats-and-tigers.html