Policy Improvement in Cribbage

Show full item record



Permalink

http://urn.fi/URN:NBN:fi-fe201804208656
Title: Policy Improvement in Cribbage
Author: Lang, Sean Ryan
Contributor: University of Helsinki, Faculty of Science, Department of Computer Science
Publisher: Helsingin yliopisto
Date: 2018
Language: eng
URI: http://urn.fi/URN:NBN:fi-fe201804208656
http://hdl.handle.net/10138/273545
Thesis level: master's thesis
Abstract: Cribbage is a card game involving multiple methods of scoring which each receive varying emphasis over the course of a typical game. Reinforcement learning is a machine learning strategy in which an agent learns to accomplish a task via direct experience by collecting rewards based on performance. In this thesis, reinforcement learning is applied to the game of cribbage, improving an agent’s policy of combining multiple basic strategies, according to the needs of the dynamic state of the game. From inspection, a reasonable policy is learned by the agent over the course of a million games, but an increase in performance was not demonstrated.


Files in this item

Total number of downloads: Loading...

Files Size Format View
SeanLang_MS_thesis.pdf 4.463Mb PDF View/Open

This item appears in the following Collection(s)

Show full item record