Poster (Scientific congresses and symposiums)
Biorthogonalization Techniques for Least Squares Temporal Difference Learning
Jung, Tobias; Ernst, Damien
2012Neural Information Processing Systems (NIPS)
 

Files


Full Text
NIPS12poster.pdf
Author postprint (156.04 kB)
DIN A0
Download
Annexes
main.pdf
Publisher postprint (250.81 kB)
Extended abstract (4 pages) (Work in progess)
Request a copy

All documents in ORBi are protected by a user license.

Send to



Details



Keywords :
Approximate dynamic programming; feature selection; sparsity; approximation
Abstract :
[en] We consider Markov reward processes and study OLS-LSTD, a framework for selecting basis functions from a set of candidates to obtain a sparse representation of the value function in the context of least squares temporal difference learning. To support efficient both updating and downdating operations, OLS-LSTD uses a biorthogonal representation for the selected basis vectors. Empirical comparisons with the recently proposed MP and LARS frameworks for LSTD are made.
Disciplines :
Computer science
Author, co-author :
Jung, Tobias ;  Université de Liège - ULiège > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Réseaux informatiques
Ernst, Damien  ;  Université de Liège - ULiège > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Smart grids
Language :
English
Title :
Biorthogonalization Techniques for Least Squares Temporal Difference Learning
Publication date :
07 December 2012
Number of pages :
A0
Event name :
Neural Information Processing Systems (NIPS)
Event organizer :
NIPS Foundation
Event place :
South Lake Tahoe, NV, United States
Event date :
from 3-12-2012 to 8-12-2012
Audience :
International
Available on ORBi :
since 14 December 2012

Statistics


Number of views
90 (17 by ULiège)
Number of downloads
76 (9 by ULiège)

Bibliography


Similar publications



Contact ORBi