Feel free to watch our github for the latest version of our program. Note that our program require the support of Keras (1.0+) to run properly.

Final Report

In this paper, we will describe the design of AlphaGo and how our program adopted various strategies of it. We will also discuss the adaptations made to fit another game and different challenges faced and overcome throughout the project. Lastly, we will showcase the experiments done and how the data collected can expedite the development in future works.

Intermediate Report

This intermediate report will describe the design and implementation of our program. It will then explain the main algorithms and justify the choices used in the design. Finally, it will showcase our progress of the project, which is the Graphical User Interface, game rule implementation, data collection and value network.This intermediate report serves the purpose of reporting the progress of our project thus far. As of now, simple neural network and skeleton of our program has been developed.

Playing Othello by Deep Learing Neural Network

AlphaGo has been a success in the fields of Artificial Intelligence and a huge news worldwide. My project aims at replicating this in the field of Computer Othello using similar techniques, such as Monte-Carlo Tree Search, value evaluation neural network and rollout policy. Ulitmately, the goal of this project is to create a program that can play Othello at superhuman level using less resources than current models.

Mastering The Game of Go with Deep Neural Network and Tree Search

In this article, the research team of AlphaGo explains their methodology used to create a professional level Go playing program. They also highlight the difficulties encountered. This article provides lots of insights and provides the main skeleton for this project.