Javascript must be enabled for the correct page display

Solving Chess Endgames Using Q Learning

Boar, Antonio-Ionut (2022) Solving Chess Endgames Using Q Learning. Bachelor's Thesis, Artificial Intelligence.


Download (666kB) | Preview
[img] Text
Restricted to Registered users only

Download (119kB)


Chess is one of the oldest and consistently popular games in human history. The same is reflected in computing history, with attempts getting better and better at developing machines that can play the game. However, most of the academic literature centers around more complex solutions than simple Reinforcement Learning. Therefore, this research reduces the scope and complexity, aiming to explore the ability of Reinforcement Learning algorithms, Q Learning specifically, to learn how to checkmate in a winning endgame scenario. This paper shows that a Q Learning agent can yield results that are worse, but comparable to more advanced chess engines in the well-known endgames of King and Queen, King and Rook and King and Two Bishops.

Item Type: Thesis (Bachelor's Thesis)
Supervisor name: Sabatelli, M.
Degree programme: Artificial Intelligence
Thesis type: Bachelor's Thesis
Language: English
Date Deposited: 16 Aug 2022 11:21
Last Modified: 16 Aug 2022 11:21

Actions (login required)

View Item View Item