Multi-agent Deep Reinforcement Learning for Task Allocation in Dynamic Environment

Dhouha Ben Noureddine, Atef Gharbi, Samir Ben Ahmed

Abstract

The task allocation problem in a distributed environment is one of the most challenging problems in a multiagent system. We propose a new task allocation process using deep reinforcement learning that allows cooperating agents to act automatically and learn how to communicate with other neighboring agents to allocate tasks and share resources. Through learning capabilities, agents will be able to reason conveniently, generate an appropriate policy and make a good decision. Our experiments show that it is possible to allocate tasks using deep Q-learning and more importantly show the performance of our distributed task allocation approach.

Download


Paper Citation


in Harvard Style

Ben Noureddine D., Gharbi A. and Ben Ahmed S. (2017). Multi-agent Deep Reinforcement Learning for Task Allocation in Dynamic Environment . In Proceedings of the 12th International Conference on Software Technologies - Volume 1: ICSOFT, ISBN 978-989-758-262-2, pages 17-26. DOI: 10.5220/0006393400170026


in Bibtex Style

@conference{icsoft17,
author={Dhouha Ben Noureddine and Atef Gharbi and Samir Ben Ahmed},
title={Multi-agent Deep Reinforcement Learning for Task Allocation in Dynamic Environment},
booktitle={Proceedings of the 12th International Conference on Software Technologies - Volume 1: ICSOFT,},
year={2017},
pages={17-26},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0006393400170026},
isbn={978-989-758-262-2},
}


in EndNote Style

TY - CONF
JO - Proceedings of the 12th International Conference on Software Technologies - Volume 1: ICSOFT,
TI - Multi-agent Deep Reinforcement Learning for Task Allocation in Dynamic Environment
SN - 978-989-758-262-2
AU - Ben Noureddine D.
AU - Gharbi A.
AU - Ben Ahmed S.
PY - 2017
SP - 17
EP - 26
DO - 10.5220/0006393400170026