A comparative study of policies in Q-learning for foraging tasks

Yogeswaran Mohan, S. G. Ponnambalam, Jawaid Iqbal Inayat Hussain

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Citations (Scopus)

Abstract

Q-learning is a machine learning technique that learns what to do and how to map states to actions to maximize rewards. Q-learning has been applied to various tasks such as foraging, soccer and prey-pursuing robots. In this paper, a simple foraging task has been considered to study the influences of the policies reported in the open literatures. A mobile robot is used to search and retrieve pucks back to a home location. The goal of this study is to identify an efficient policy for q-learning which maximizes the number of pucks collected and minimizes the number of collisions in the environment. Policies namely greedy, epsilon-greedy, Boltzmann distribution and random search are used to study their performances in the foraging task and the results are presented.

Original languageEnglish
Title of host publication2009 World Congress on Nature and Biologically Inspired Computing, NABIC 2009 - Proceedings
Pages134-139
Number of pages6
DOIs
Publication statusPublished - 01 Dec 2009
Event2009 World Congress on Nature and Biologically Inspired Computing, NABIC 2009 - Coimbatore, India
Duration: 09 Dec 200911 Dec 2009

Publication series

Name2009 World Congress on Nature and Biologically Inspired Computing, NABIC 2009 - Proceedings

Other

Other2009 World Congress on Nature and Biologically Inspired Computing, NABIC 2009
CountryIndia
CityCoimbatore
Period09/12/0911/12/09

Fingerprint

Mobile robots
Learning systems
Robots

All Science Journal Classification (ASJC) codes

  • Artificial Intelligence
  • Computer Science Applications
  • Software

Cite this

Mohan, Y., Ponnambalam, S. G., & Inayat Hussain, J. I. (2009). A comparative study of policies in Q-learning for foraging tasks. In 2009 World Congress on Nature and Biologically Inspired Computing, NABIC 2009 - Proceedings (pp. 134-139). [5393616] (2009 World Congress on Nature and Biologically Inspired Computing, NABIC 2009 - Proceedings). https://doi.org/10.1109/NABIC.2009.5393616
Mohan, Yogeswaran ; Ponnambalam, S. G. ; Inayat Hussain, Jawaid Iqbal. / A comparative study of policies in Q-learning for foraging tasks. 2009 World Congress on Nature and Biologically Inspired Computing, NABIC 2009 - Proceedings. 2009. pp. 134-139 (2009 World Congress on Nature and Biologically Inspired Computing, NABIC 2009 - Proceedings).
@inproceedings{41cc74470e8948f5949b1edf43bc4e48,
title = "A comparative study of policies in Q-learning for foraging tasks",
abstract = "Q-learning is a machine learning technique that learns what to do and how to map states to actions to maximize rewards. Q-learning has been applied to various tasks such as foraging, soccer and prey-pursuing robots. In this paper, a simple foraging task has been considered to study the influences of the policies reported in the open literatures. A mobile robot is used to search and retrieve pucks back to a home location. The goal of this study is to identify an efficient policy for q-learning which maximizes the number of pucks collected and minimizes the number of collisions in the environment. Policies namely greedy, epsilon-greedy, Boltzmann distribution and random search are used to study their performances in the foraging task and the results are presented.",
author = "Yogeswaran Mohan and Ponnambalam, {S. G.} and {Inayat Hussain}, {Jawaid Iqbal}",
year = "2009",
month = "12",
day = "1",
doi = "10.1109/NABIC.2009.5393616",
language = "English",
isbn = "9781424456123",
series = "2009 World Congress on Nature and Biologically Inspired Computing, NABIC 2009 - Proceedings",
pages = "134--139",
booktitle = "2009 World Congress on Nature and Biologically Inspired Computing, NABIC 2009 - Proceedings",

}

Mohan, Y, Ponnambalam, SG & Inayat Hussain, JI 2009, A comparative study of policies in Q-learning for foraging tasks. in 2009 World Congress on Nature and Biologically Inspired Computing, NABIC 2009 - Proceedings., 5393616, 2009 World Congress on Nature and Biologically Inspired Computing, NABIC 2009 - Proceedings, pp. 134-139, 2009 World Congress on Nature and Biologically Inspired Computing, NABIC 2009, Coimbatore, India, 09/12/09. https://doi.org/10.1109/NABIC.2009.5393616

A comparative study of policies in Q-learning for foraging tasks. / Mohan, Yogeswaran; Ponnambalam, S. G.; Inayat Hussain, Jawaid Iqbal.

2009 World Congress on Nature and Biologically Inspired Computing, NABIC 2009 - Proceedings. 2009. p. 134-139 5393616 (2009 World Congress on Nature and Biologically Inspired Computing, NABIC 2009 - Proceedings).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - A comparative study of policies in Q-learning for foraging tasks

AU - Mohan, Yogeswaran

AU - Ponnambalam, S. G.

AU - Inayat Hussain, Jawaid Iqbal

PY - 2009/12/1

Y1 - 2009/12/1

N2 - Q-learning is a machine learning technique that learns what to do and how to map states to actions to maximize rewards. Q-learning has been applied to various tasks such as foraging, soccer and prey-pursuing robots. In this paper, a simple foraging task has been considered to study the influences of the policies reported in the open literatures. A mobile robot is used to search and retrieve pucks back to a home location. The goal of this study is to identify an efficient policy for q-learning which maximizes the number of pucks collected and minimizes the number of collisions in the environment. Policies namely greedy, epsilon-greedy, Boltzmann distribution and random search are used to study their performances in the foraging task and the results are presented.

AB - Q-learning is a machine learning technique that learns what to do and how to map states to actions to maximize rewards. Q-learning has been applied to various tasks such as foraging, soccer and prey-pursuing robots. In this paper, a simple foraging task has been considered to study the influences of the policies reported in the open literatures. A mobile robot is used to search and retrieve pucks back to a home location. The goal of this study is to identify an efficient policy for q-learning which maximizes the number of pucks collected and minimizes the number of collisions in the environment. Policies namely greedy, epsilon-greedy, Boltzmann distribution and random search are used to study their performances in the foraging task and the results are presented.

UR - http://www.scopus.com/inward/record.url?scp=77949575197&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77949575197&partnerID=8YFLogxK

U2 - 10.1109/NABIC.2009.5393616

DO - 10.1109/NABIC.2009.5393616

M3 - Conference contribution

SN - 9781424456123

T3 - 2009 World Congress on Nature and Biologically Inspired Computing, NABIC 2009 - Proceedings

SP - 134

EP - 139

BT - 2009 World Congress on Nature and Biologically Inspired Computing, NABIC 2009 - Proceedings

ER -

Mohan Y, Ponnambalam SG, Inayat Hussain JI. A comparative study of policies in Q-learning for foraging tasks. In 2009 World Congress on Nature and Biologically Inspired Computing, NABIC 2009 - Proceedings. 2009. p. 134-139. 5393616. (2009 World Congress on Nature and Biologically Inspired Computing, NABIC 2009 - Proceedings). https://doi.org/10.1109/NABIC.2009.5393616