Combining Monte Carlo tree search and apprenticeship learning for capture the flag

Ivanovo, J; Raffe, WL; Zambetta, F; Li, X

Combining Monte Carlo tree search and apprenticeship learning for capture the flag

Ivanovo, J Raffe, WL

Zambetta, F Li, X

Permalink

Publication Type:: Conference Proceeding
Citation:: 2015 IEEE Conference on Computational Intelligence and Games, CIG 2015 - Proceedings, 2015, pp. 154 - 161
Issue Date:: 2015-11-04

Closed Access

	Filename	Description	Size
	07317914.pdf	Published version	1.01 MB	Adobe PDF	View/Open

Copyright Clearance Process

Recently Added
In Progress
Closed Access

This item is closed access and not available.

Full metadata record

Field	Value	Language
dc.contributor.author	Ivanovo, J	en_US
dc.contributor.author	Raffe, WL https://orcid.org/0000-0001-5310-0943	en_US
dc.contributor.author	Zambetta, F	en_US
dc.contributor.author	Li, X	en_US
dc.date.issued	2015-11-04	en_US
dc.identifier.citation	2015 IEEE Conference on Computational Intelligence and Games, CIG 2015 - Proceedings, 2015, pp. 154 - 161	en_US
dc.identifier.isbn	9781479986217	en_US
dc.identifier.uri	http://hdl.handle.net/10453/119864
dc.description.abstract	© 2015 IEEE. In this paper we introduce a novel approach to agent control in competitive video games which combines Monte Carlo Tree Search (MCTS) and Apprenticeship Learning (AL). More specifically, an opponent model created through AL is used during the expansion phase of the Upper Confidence Bounds for Trees (UCT) variant of MCTS. We show how this approach can be applied to a game of Capture the Flag (CTF), an environment which is both non-deterministic and partially observable. The performance gain of a controller utilizing an opponent model learned via AL when compared to a controller using just UCT is shown both with win/loss ratios and True Skill rankings. Additionally, we build on previous findings by providing evidence of a bias towards a particular style of play in the AI Sandbox CTF environment. We believe that the approach highlighted here can be extended to a wider range of games other than just CTF.	en_US
dc.relation.ispartof	2015 IEEE Conference on Computational Intelligence and Games, CIG 2015 - Proceedings	en_US
dc.relation.isbasedon	10.1109/CIG.2015.7317914	en_US
dc.title	Combining Monte Carlo tree search and apprenticeship learning for capture the flag	en_US
dc.type	Conference Proceeding
utslib.for	0806 Information Systems	en_US
pubs.embargo.period	Not known	en_US
pubs.organisational-group	/University of Technology Sydney
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology/School of Computer Science
pubs.organisational-group	/University of Technology Sydney/Strength - PERSWADE - Centre on Persuasive Systems for Wise Adaptive Living
utslib.copyright.status	closed_access
pubs.publication-status	Published	en_US

Abstract:

© 2015 IEEE. In this paper we introduce a novel approach to agent control in competitive video games which combines Monte Carlo Tree Search (MCTS) and Apprenticeship Learning (AL). More specifically, an opponent model created through AL is used during the expansion phase of the Upper Confidence Bounds for Trees (UCT) variant of MCTS. We show how this approach can be applied to a game of Capture the Flag (CTF), an environment which is both non-deterministic and partially observable. The performance gain of a controller utilizing an opponent model learned via AL when compared to a controller using just UCT is shown both with win/loss ratios and True Skill rankings. Additionally, we build on previous findings by providing evidence of a bias towards a particular style of play in the AI Sandbox CTF environment. We believe that the approach highlighted here can be extended to a wider range of games other than just CTF.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/119864