Introducing communication in Dis-POMDPs with locality of interaction

Makoto Tasaki, Yuichi Yabu, Yuki Iwanari, Makoto Yokoo, Milind Tambe, Janusz Marecki, Pradeep Varakantham

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

4 Citations (Scopus)

Abstract

The Networked Distributed POMDPs (ND-POMDPs) can model multiagent systems in uncertain domains and has begun to scale-up the number of agents. However, prior work in ND-POMDPs has failed to address communication. Without communication, the size of a local policy at each agent within the ND-POMDPs grows exponentially in the time horizon. To overcome this problem, we extend existing algorithms so that agents periodically communicate their observation and action histories with each other. After communication, agents can start from new synchronized belief state. Thus, we can avoid the exponential growth in the size of local policies at agents. Furthermore, we introduce an idea that is similar the Point-based Value Iteration algorithm to approximate the value function with a fixed number of representative points. Our experimental results show that we can obtain much longer policies than existing algorithms as long as the interval between communications is small.

Original languageEnglish
Title of host publicationProceedings - 2008 IEEE/WIC/ACM International Conference on Intelligent Agent Technology, IAT 2008
Pages169-175
Number of pages7
DOIs
Publication statusPublished - Dec 1 2008
Event2008 IEEE/WIC/ACM International Conference on Intelligent Agent Technology, IAT 2008 - Sydney, NSW, Australia
Duration: Dec 9 2008Dec 12 2008

Publication series

NameProceedings - 2008 IEEE/WIC/ACM International Conference on Intelligent Agent Technology, IAT 2008

Other

Other2008 IEEE/WIC/ACM International Conference on Intelligent Agent Technology, IAT 2008
CountryAustralia
CitySydney, NSW
Period12/9/0812/12/08

Fingerprint

Communication
Multi agent systems

All Science Journal Classification (ASJC) codes

  • Artificial Intelligence
  • Software

Cite this

Tasaki, M., Yabu, Y., Iwanari, Y., Yokoo, M., Tambe, M., Marecki, J., & Varakantham, P. (2008). Introducing communication in Dis-POMDPs with locality of interaction. In Proceedings - 2008 IEEE/WIC/ACM International Conference on Intelligent Agent Technology, IAT 2008 (pp. 169-175). [4740618] (Proceedings - 2008 IEEE/WIC/ACM International Conference on Intelligent Agent Technology, IAT 2008). https://doi.org/10.1109/WIIAT.2008.316

Introducing communication in Dis-POMDPs with locality of interaction. / Tasaki, Makoto; Yabu, Yuichi; Iwanari, Yuki; Yokoo, Makoto; Tambe, Milind; Marecki, Janusz; Varakantham, Pradeep.

Proceedings - 2008 IEEE/WIC/ACM International Conference on Intelligent Agent Technology, IAT 2008. 2008. p. 169-175 4740618 (Proceedings - 2008 IEEE/WIC/ACM International Conference on Intelligent Agent Technology, IAT 2008).

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

Tasaki, M, Yabu, Y, Iwanari, Y, Yokoo, M, Tambe, M, Marecki, J & Varakantham, P 2008, Introducing communication in Dis-POMDPs with locality of interaction. in Proceedings - 2008 IEEE/WIC/ACM International Conference on Intelligent Agent Technology, IAT 2008., 4740618, Proceedings - 2008 IEEE/WIC/ACM International Conference on Intelligent Agent Technology, IAT 2008, pp. 169-175, 2008 IEEE/WIC/ACM International Conference on Intelligent Agent Technology, IAT 2008, Sydney, NSW, Australia, 12/9/08. https://doi.org/10.1109/WIIAT.2008.316
Tasaki M, Yabu Y, Iwanari Y, Yokoo M, Tambe M, Marecki J et al. Introducing communication in Dis-POMDPs with locality of interaction. In Proceedings - 2008 IEEE/WIC/ACM International Conference on Intelligent Agent Technology, IAT 2008. 2008. p. 169-175. 4740618. (Proceedings - 2008 IEEE/WIC/ACM International Conference on Intelligent Agent Technology, IAT 2008). https://doi.org/10.1109/WIIAT.2008.316
Tasaki, Makoto ; Yabu, Yuichi ; Iwanari, Yuki ; Yokoo, Makoto ; Tambe, Milind ; Marecki, Janusz ; Varakantham, Pradeep. / Introducing communication in Dis-POMDPs with locality of interaction. Proceedings - 2008 IEEE/WIC/ACM International Conference on Intelligent Agent Technology, IAT 2008. 2008. pp. 169-175 (Proceedings - 2008 IEEE/WIC/ACM International Conference on Intelligent Agent Technology, IAT 2008).
@inproceedings{f9cf37772da84299abbb6da171ac21b5,
title = "Introducing communication in Dis-POMDPs with locality of interaction",
abstract = "The Networked Distributed POMDPs (ND-POMDPs) can model multiagent systems in uncertain domains and has begun to scale-up the number of agents. However, prior work in ND-POMDPs has failed to address communication. Without communication, the size of a local policy at each agent within the ND-POMDPs grows exponentially in the time horizon. To overcome this problem, we extend existing algorithms so that agents periodically communicate their observation and action histories with each other. After communication, agents can start from new synchronized belief state. Thus, we can avoid the exponential growth in the size of local policies at agents. Furthermore, we introduce an idea that is similar the Point-based Value Iteration algorithm to approximate the value function with a fixed number of representative points. Our experimental results show that we can obtain much longer policies than existing algorithms as long as the interval between communications is small.",
author = "Makoto Tasaki and Yuichi Yabu and Yuki Iwanari and Makoto Yokoo and Milind Tambe and Janusz Marecki and Pradeep Varakantham",
year = "2008",
month = "12",
day = "1",
doi = "10.1109/WIIAT.2008.316",
language = "English",
isbn = "9780769534961",
series = "Proceedings - 2008 IEEE/WIC/ACM International Conference on Intelligent Agent Technology, IAT 2008",
pages = "169--175",
booktitle = "Proceedings - 2008 IEEE/WIC/ACM International Conference on Intelligent Agent Technology, IAT 2008",

}

TY - GEN

T1 - Introducing communication in Dis-POMDPs with locality of interaction

AU - Tasaki, Makoto

AU - Yabu, Yuichi

AU - Iwanari, Yuki

AU - Yokoo, Makoto

AU - Tambe, Milind

AU - Marecki, Janusz

AU - Varakantham, Pradeep

PY - 2008/12/1

Y1 - 2008/12/1

N2 - The Networked Distributed POMDPs (ND-POMDPs) can model multiagent systems in uncertain domains and has begun to scale-up the number of agents. However, prior work in ND-POMDPs has failed to address communication. Without communication, the size of a local policy at each agent within the ND-POMDPs grows exponentially in the time horizon. To overcome this problem, we extend existing algorithms so that agents periodically communicate their observation and action histories with each other. After communication, agents can start from new synchronized belief state. Thus, we can avoid the exponential growth in the size of local policies at agents. Furthermore, we introduce an idea that is similar the Point-based Value Iteration algorithm to approximate the value function with a fixed number of representative points. Our experimental results show that we can obtain much longer policies than existing algorithms as long as the interval between communications is small.

AB - The Networked Distributed POMDPs (ND-POMDPs) can model multiagent systems in uncertain domains and has begun to scale-up the number of agents. However, prior work in ND-POMDPs has failed to address communication. Without communication, the size of a local policy at each agent within the ND-POMDPs grows exponentially in the time horizon. To overcome this problem, we extend existing algorithms so that agents periodically communicate their observation and action histories with each other. After communication, agents can start from new synchronized belief state. Thus, we can avoid the exponential growth in the size of local policies at agents. Furthermore, we introduce an idea that is similar the Point-based Value Iteration algorithm to approximate the value function with a fixed number of representative points. Our experimental results show that we can obtain much longer policies than existing algorithms as long as the interval between communications is small.

UR - http://www.scopus.com/inward/record.url?scp=62949185084&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=62949185084&partnerID=8YFLogxK

U2 - 10.1109/WIIAT.2008.316

DO - 10.1109/WIIAT.2008.316

M3 - Conference contribution

AN - SCOPUS:62949185084

SN - 9780769534961

T3 - Proceedings - 2008 IEEE/WIC/ACM International Conference on Intelligent Agent Technology, IAT 2008

SP - 169

EP - 175

BT - Proceedings - 2008 IEEE/WIC/ACM International Conference on Intelligent Agent Technology, IAT 2008

ER -