Online Sensor Selection in Reinforcement Learning Environment

Koichiro Ishikawa, Tsutomu Fujinami, Susumu Kunifuji, Akito Sakurai

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

More sensors do not necessarily result in more appropriate state descriptions, so that a mobile robot has to select an appropriate set of sensors besides learning a state-action function in a reinforcement learning environment. We present a multi-armed bandit formulation of the problem and apply it to mobile robot navigation task. We modified the reinforcement comparison method to suit our problem and build a system where the selection of optimal set of sensors and the learning of state-action functions are done simultaneously. Our approach is evaluated on a Khepera robot simulator and the results reveal that our approach works well as an integrated learning system to identify the best set of sensors and reduce learning time.

Original languageEnglish
Pages (from-to)870-878
Number of pages9
JournalIEEJ Transactions on Electronics, Information and Systems
Volume125
Issue number6
DOIs
Publication statusPublished - 2005

Fingerprint

Reinforcement learning
Sensors
Mobile robots
Learning systems
Reinforcement
Navigation
Simulators
Robots

Keywords

  • autonomous mobile robot
  • Q-learning
  • reinforcement learning
  • sensor selection

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Cite this

Online Sensor Selection in Reinforcement Learning Environment. / Ishikawa, Koichiro; Fujinami, Tsutomu; Kunifuji, Susumu; Sakurai, Akito.

In: IEEJ Transactions on Electronics, Information and Systems, Vol. 125, No. 6, 2005, p. 870-878.

Research output: Contribution to journalArticle

Ishikawa, Koichiro ; Fujinami, Tsutomu ; Kunifuji, Susumu ; Sakurai, Akito. / Online Sensor Selection in Reinforcement Learning Environment. In: IEEJ Transactions on Electronics, Information and Systems. 2005 ; Vol. 125, No. 6. pp. 870-878.
@article{64bb1b6444c448589b74a4bce803cea2,
title = "Online Sensor Selection in Reinforcement Learning Environment",
abstract = "More sensors do not necessarily result in more appropriate state descriptions, so that a mobile robot has to select an appropriate set of sensors besides learning a state-action function in a reinforcement learning environment. We present a multi-armed bandit formulation of the problem and apply it to mobile robot navigation task. We modified the reinforcement comparison method to suit our problem and build a system where the selection of optimal set of sensors and the learning of state-action functions are done simultaneously. Our approach is evaluated on a Khepera robot simulator and the results reveal that our approach works well as an integrated learning system to identify the best set of sensors and reduce learning time.",
keywords = "autonomous mobile robot, Q-learning, reinforcement learning, sensor selection",
author = "Koichiro Ishikawa and Tsutomu Fujinami and Susumu Kunifuji and Akito Sakurai",
year = "2005",
doi = "10.1541/ieejeiss.125.870",
language = "English",
volume = "125",
pages = "870--878",
journal = "IEEJ Transactions on Electronics, Information and Systems",
issn = "0385-4221",
publisher = "The Institute of Electrical Engineers of Japan",
number = "6",

}

TY - JOUR

T1 - Online Sensor Selection in Reinforcement Learning Environment

AU - Ishikawa, Koichiro

AU - Fujinami, Tsutomu

AU - Kunifuji, Susumu

AU - Sakurai, Akito

PY - 2005

Y1 - 2005

N2 - More sensors do not necessarily result in more appropriate state descriptions, so that a mobile robot has to select an appropriate set of sensors besides learning a state-action function in a reinforcement learning environment. We present a multi-armed bandit formulation of the problem and apply it to mobile robot navigation task. We modified the reinforcement comparison method to suit our problem and build a system where the selection of optimal set of sensors and the learning of state-action functions are done simultaneously. Our approach is evaluated on a Khepera robot simulator and the results reveal that our approach works well as an integrated learning system to identify the best set of sensors and reduce learning time.

AB - More sensors do not necessarily result in more appropriate state descriptions, so that a mobile robot has to select an appropriate set of sensors besides learning a state-action function in a reinforcement learning environment. We present a multi-armed bandit formulation of the problem and apply it to mobile robot navigation task. We modified the reinforcement comparison method to suit our problem and build a system where the selection of optimal set of sensors and the learning of state-action functions are done simultaneously. Our approach is evaluated on a Khepera robot simulator and the results reveal that our approach works well as an integrated learning system to identify the best set of sensors and reduce learning time.

KW - autonomous mobile robot

KW - Q-learning

KW - reinforcement learning

KW - sensor selection

UR - http://www.scopus.com/inward/record.url?scp=33749550931&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33749550931&partnerID=8YFLogxK

U2 - 10.1541/ieejeiss.125.870

DO - 10.1541/ieejeiss.125.870

M3 - Article

AN - SCOPUS:33749550931

VL - 125

SP - 870

EP - 878

JO - IEEJ Transactions on Electronics, Information and Systems

JF - IEEJ Transactions on Electronics, Information and Systems

SN - 0385-4221

IS - 6

ER -