Upper Confidence Trees and Billiards for Optimal Active Learning

Philippe Rolet 1, 2 Michèle Sebag 1, 2 Olivier Teytaud 1, 2
2 TAO - Machine Learning and Optimisation
CNRS - Centre National de la Recherche Scientifique : UMR8623, Inria Saclay - Ile de France, UP11 - Université Paris-Sud - Paris 11, LRI - Laboratoire de Recherche en Informatique
Abstract : This paper focuses on Active Learning (AL) with bounded compu- tational resources. AL is formalized as a finite horizon Reinforcement Learning problem, and tackled as a single-player game. An approximate optimal AL strat- egy based on tree-structured multi-armed bandit algorithms and billiard-based sampling is presented together with a proof of principle of the approach.
Document type :
Conference papers
Complete list of metadatas

Cited literature [19 references]  Display  Hide  Download

https://hal.inria.fr/inria-00369787
Contributor : Olivier Teytaud <>
Submitted on : Saturday, March 21, 2009 - 9:58:28 AM
Last modification on : Thursday, April 5, 2018 - 12:30:12 PM
Long-term archiving on : Friday, October 12, 2012 - 2:05:35 PM

File

capfinal.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : inria-00369787, version 1

Collections

Citation

Philippe Rolet, Michèle Sebag, Olivier Teytaud. Upper Confidence Trees and Billiards for Optimal Active Learning. CAP09, 2009, Hammamet, Tunisie, Tunisia. ⟨inria-00369787⟩

Share

Metrics

Record views

413

Files downloads

292