Document Detail

An information-theoretic analysis of return maximization in reinforcement learning.
MedLine Citation:
PMID:  21665429     Owner:  NLM     Status:  Publisher    
We present a general analysis of return maximization in reinforcement learning. This analysis does not require assumptions of Markovianity, stationarity, and ergodicity for the stochastic sequential decision processes of reinforcement learning. Instead, our analysis assumes the asymptotic equipartition property fundamental to information theory, providing a substantially different view from that in the literature. As our main results, we show that return maximization is achieved by the overlap of typical and best sequence sets, and we present a class of stochastic sequential decision processes with the necessary condition for return maximization. We also describe several examples of best sequences in terms of return maximization in the class of stochastic sequential decision processes, which satisfy the necessary condition.
Kazunori Iwata
Related Documents :
21840409 - Multi-task functional mri in multiple sclerosis patients without clinical disability.
21562269 - The activity-regulated cytoskeletal-associated protein (arc/arg3.1) is required for rec...
21987919 - Evaluation of task performance during mentally imaging three-dimensional shapes from pl...
22005749 - Ampa receptor endocytosis in rat perirhinal cortex underlies retrieval of object memory.
21689669 - Executive function mediates effects of white matter hyperintensities on episodic memory.
21995709 - Using sensecam with an amnesic patient: accessing inaccessible everyday memories.
21962459 - Early cognitive profiles of emergent readers: a longitudinal study.
20926099 - Connecting instances to promote children's relational reasoning.
19045799 - Current-level discrimination and spectral profile analysis in multi-channel electrical ...
Publication Detail:
Type:  JOURNAL ARTICLE     Date:  2011-5-17
Journal Detail:
Title:  Neural networks : the official journal of the International Neural Network Society     Volume:  -     ISSN:  1879-2782     ISO Abbreviation:  -     Publication Date:  2011 May 
Date Detail:
Created Date:  2011-6-13     Completed Date:  -     Revised Date:  -    
Medline Journal Info:
Nlm Unique ID:  8805018     Medline TA:  Neural Netw     Country:  -    
Other Details:
Languages:  ENG     Pagination:  -     Citation Subset:  -    
Copyright Information:
Copyright © 2011 Elsevier Ltd. All rights reserved.
Export Citation:
APA/MLA Format     Download EndNote     Download BibTex
MeSH Terms

From MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine

Previous Document:  On the road to invariant recognition: Explaining tradeoff and morph properties of cells in inferotem...
Next Document:  Deterioration of Myocardial Function in Paradoxical Low-Flow Severe Aortic Stenosis: Two-Dimensional...