Bachelor Thesis: Chess App

This thesis is about creating a web application that allows users to play chess against a learning chess engine after each game.

The project is divided into two parts.

The initial stage entails creating a chess game architecture upon which the web application is constructed.

The web application will have a modern, curated user interface via which users will be able to interact with the chess board and make moves. Additional information, such as captured pieces and a move history, are also provided through the interface. To locate solid moves, the user can also employ extra tools such as rolling back movements or receiving assistance from an integrated third-party chess engine.

The second portion of the thesis involves developing a chess engine that will compete with users. Following each game, the engine will study plays and change its parameters to improve future gaming strategies. This means that the more games are played against the engine, the better it will become. The engine will be educated using a reinforcement learning technique, which allows it to improve and learn from past games’ results. To track development, the engine will be compared to an existing third-party engine to determine an estimate of its own Elo rating.

This thesis aims to develop a chess online application with an engine that continuously grows its own Elo after each game, at least until it achieves the Elo of an average chess player, and so beats the average person.

Student: 1
Contact: philipp.zech@uibk.ac.at