Rochester Institute of Technology
Using Machine Learning Algorithms to Predict Outcomes of Chess Games Using Player Data
Abstract
dc:description.abstractChess is a two-player game, popular with a wide variety of people, ranging from people who play casually every once in a while with their family to professional players who make their living through playing and teaching. There are three outcomes that a chess game could have: player 1 (White) wins, player 2 (Black) wins, or both players draw. Using an online database called “The Week In Chess” which contains information about 2.5 million chess games, a variety of machine learning methods are tested to predict the outcomes of chess games. The features investigated as inputs for classification are: White’s Win Rate, White’s Draw Rate, Black’s Win Rate, Black’s’ Draw Rate, Opening’s Win Rate and Opening’s Draw Rate. Then, the results of the classifiers using the full set of features and various subsets of features are compared in order to try and determine the best possible accuracy in classifying the game outcome.
Degree
thesis:*- Name thesis:degree_name
- Applied and Computational Mathematics (MS)
- Level thesis:degree_level
- Thesis
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- DeCredico, Sofia
- Contributors dc:contributor
-
- Nathan Cahill
Subjects
dc:subject × 1Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.rit.edu/theses/11905