Enhanced temporal difference learning using compiled eligibility traces

Title: Enhanced temporal difference learning using compiled eligibility traces
Creator: Vamplew, Peter; Ollington, Robert; Hepburn, Mark
Date: 2006
Type: Text; Conference paper
Identifier: http://researchonline.federation.edu.au/vital/access/HandleResolver/1959.17/41384
Identifier: vital:1610
Identifier: https://doi.org/10.1007/11941439_18
Abstract: Eligibility traces have been shown to substantially improve the convergence speed of temporal difference learning algorithms, by maintaining a record of recently experienced states. This paper presents an extension of conventional eligibility traces (compiled traces) which retain additional information about the agent’s experience within the environment. Empirical results show that compiled traces outperform conventional traces when applied to policy evaluation tasks using a tabular representation of the state values.; E1
Publisher: Hobart, Australia : Springer
Relation: Paper presented at Artificial Intelligence, AI 2006: Advances in Artificial Intelligence conference, Hobart, Australia : 4th December, 2006
Rights: Copyright Springer
Rights: This metadata is freely available under a CCO license
Subject: Learning; Teaching; Temporal difference
Reviewed

		Thumbnail	File	Description	Size	Format