Evaluating Go Game Records for Prediction of Player Attributes

0459376 - ÚI 2017 RIV US eng C - Conference Paper (international conference)
Moudřík, J. - Baudiš, P. - Neruda, Roman
Evaluating Go Game Records for Prediction of Player Attributes.
Proceedings of the 2015 IEEE Conference on Computational Intelligence and Games. Piscataway: IEEE, 2015, s. 162-168. ISBN 978-1-4799-8622-4. ISSN 2325-4289.
[CIG 2015. IEEE Conference on Computational Intelligence and Games. Tainan (TW), 31.08.2015-05.09.2015]
R&D Projects: GA ČR GA15-19877S
Grant - others:GA UK(CZ) 364015; SVV(CZ) 260 224
Institutional support: RVO:67985807
Keywords : computer Go * machine learning * feature extraction * board games * skill assessment
Subject RIV: IN - Informatics, Computer Science

We propose a way of extracting and aggregating permove evaluations from sets of Go game records. The evaluations capture different aspects of the games such as played patterns or statistic of sente/gote sequences. Using machine learning algorithms, the evaluations can be utilized to predict different relevant target variables. We apply this methodology to predict the strength and playing style of the player (e.g. territoriality or aggressivity) with good accuracy. We propose a number of possible applications including aiding in Go study, seeding realwork ranks of internet players or tuning of Go-playing programs.
Permanent Link: http://hdl.handle.net/11104/0259581