Selecting Near-Optimal Approximate State Representations in Reinforcement Learning

Ronald Ortner 1 Odalric-Ambrym Maillard 2 Daniil Ryabko 3
1 MUL - Montanuniversität Leoben
2 EE-Technion - Department of Electrical Engineering - Technion [Haïfa]
3 SEQUEL - Sequential Learning
LIFL - Laboratoire d'Informatique Fondamentale de Lille, LAGIS - Laboratoire d'Automatique, Génie Informatique et Signal, Inria Lille - Nord Europe
Abstract : We consider a reinforcement learning setting where the learner does not have explicit access to the states of the underlying Markov decision process (MDP). Instead, she has access to several models that map histories of past interactions to states. Here we improve over known regret bounds in this setting, and more importantly generalize to the case where the models given to the learner do not contain a true model resulting in an MDP representation but only approximations of it. We also give improved error bounds for state aggregation.
Document type :
Conference papers
Complete list of metadatas
https://hal.inria.fr/hal-01057562
Contributor : Daniil Ryabko <>
Submitted on : Saturday, August 23, 2014 - 10:40:11 PM
Last modification on : Thursday, February 21, 2019 - 10:52:49 AM

Identifiers

  • HAL Id : hal-01057562, version 1

Collections

CNRS | INRIA | UNIV-LILLE3 | LAGIS | CRISTAL | CRISTAL-SEQUEL

Citation

Ronald Ortner, Odalric-Ambrym Maillard, Daniil Ryabko. Selecting Near-Optimal Approximate State Representations in Reinforcement Learning. International Conference on Algorithmic Learning Theory (ALT), Oct 2014, Bled, Slovenia. pp.140-154. ⟨hal-01057562⟩

Export

BibTeX TEI DC DCterms EndNote

Share

Metrics

Record views

359

 

 

Contact

archive-ouverte@inria.fr