3D Tic Tac Toe AI

By:

• Ahmed Mozammil Iqbal (ahmediq-git)
• Javeria Siddique (Javeria-Siddique)

Summary

This was our submission for our Dynamic Programming and Reinforcement Learning (CS6314) Final Project and competition.

The codebase is divided into 3 phases. The details are given in the Project Report.

The 3rd Phase was our code for the 3D Tic Tac Toe (4 x 4 x 4) Competition where we had to develop an AI or Reinforcement Learning solution that can play as a bot in the game.

We created 2 models:

• A Deep Q network
• Rule Based AI

The competition for the 3rd Phase was a tournament tnat involved all groups play their designed AI bots against the other groups AI bots in a game of 3D Tic Tac Toe.

We enlisted in our class competition with the Rule Based AI and won first position in the competition.

Special Thanks to Mahen Mughal (mahen2-cmd) for making the boiler plate code and designing the Project Requirements:

https://github.com/mahen2-cmd/Final-Project-DPRL