Learning to Improve Coordinated Actions in Cooperative Distributed Problem-Solving Environments

Toshiharu Sugawara, Victor Lesser

Research output: Contribution to journalArticle

34 Citations (Scopus)

Abstract

Coordination is an essential technique in cooperative, distributed multiagent systems. However, sophisticated coordination strategies are not always cost-effective in all problem-solving situations. This paper presents a learning method to identify what information will improve coordination in specific problem-solving situations. Learning is accomplished by recording and analyzing traces of inferences after problem solving. The analysis identifies situations where inappropriate coordination strategies caused redundant activities, or the lack of timely execution of important activities, thus degrading system performance. To remedy this problem, situation-specific control rules are created which acquire additional nonlocal information about activities in the agent networks and then select another plan or another scheduling strategy. Examples from a real distributed problem-solving application involving diagnosis of a local area network are described.

Original languageEnglish
Pages (from-to)129-153
Number of pages25
JournalMachine Learning
Volume12
Issue number4
Publication statusPublished - 1998
Externally publishedYes

Fingerprint

Problem Solving Environment
Distributed Environment
Multi agent systems
Local area networks
Multi-agent Systems
Distributed Systems
System Performance
Scheduling
Trace
Learning
Costs
Strategy

Keywords

  • Coordination
  • Learning
  • Multiagent systems

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Hardware and Architecture
  • Theoretical Computer Science
  • Computational Theory and Mathematics
  • Artificial Intelligence
  • Control and Systems Engineering

Cite this

Learning to Improve Coordinated Actions in Cooperative Distributed Problem-Solving Environments. / Sugawara, Toshiharu; Lesser, Victor.

In: Machine Learning, Vol. 12, No. 4, 1998, p. 129-153.

Research output: Contribution to journalArticle

@article{7da4e277c48f4e1083015813cdfacb04,
title = "Learning to Improve Coordinated Actions in Cooperative Distributed Problem-Solving Environments",
abstract = "Coordination is an essential technique in cooperative, distributed multiagent systems. However, sophisticated coordination strategies are not always cost-effective in all problem-solving situations. This paper presents a learning method to identify what information will improve coordination in specific problem-solving situations. Learning is accomplished by recording and analyzing traces of inferences after problem solving. The analysis identifies situations where inappropriate coordination strategies caused redundant activities, or the lack of timely execution of important activities, thus degrading system performance. To remedy this problem, situation-specific control rules are created which acquire additional nonlocal information about activities in the agent networks and then select another plan or another scheduling strategy. Examples from a real distributed problem-solving application involving diagnosis of a local area network are described.",
keywords = "Coordination, Learning, Multiagent systems",
author = "Toshiharu Sugawara and Victor Lesser",
year = "1998",
language = "English",
volume = "12",
pages = "129--153",
journal = "Machine Learning",
issn = "0885-6125",
publisher = "Springer Netherlands",
number = "4",

}

TY - JOUR

T1 - Learning to Improve Coordinated Actions in Cooperative Distributed Problem-Solving Environments

AU - Sugawara, Toshiharu

AU - Lesser, Victor

PY - 1998

Y1 - 1998

N2 - Coordination is an essential technique in cooperative, distributed multiagent systems. However, sophisticated coordination strategies are not always cost-effective in all problem-solving situations. This paper presents a learning method to identify what information will improve coordination in specific problem-solving situations. Learning is accomplished by recording and analyzing traces of inferences after problem solving. The analysis identifies situations where inappropriate coordination strategies caused redundant activities, or the lack of timely execution of important activities, thus degrading system performance. To remedy this problem, situation-specific control rules are created which acquire additional nonlocal information about activities in the agent networks and then select another plan or another scheduling strategy. Examples from a real distributed problem-solving application involving diagnosis of a local area network are described.

AB - Coordination is an essential technique in cooperative, distributed multiagent systems. However, sophisticated coordination strategies are not always cost-effective in all problem-solving situations. This paper presents a learning method to identify what information will improve coordination in specific problem-solving situations. Learning is accomplished by recording and analyzing traces of inferences after problem solving. The analysis identifies situations where inappropriate coordination strategies caused redundant activities, or the lack of timely execution of important activities, thus degrading system performance. To remedy this problem, situation-specific control rules are created which acquire additional nonlocal information about activities in the agent networks and then select another plan or another scheduling strategy. Examples from a real distributed problem-solving application involving diagnosis of a local area network are described.

KW - Coordination

KW - Learning

KW - Multiagent systems

UR - http://www.scopus.com/inward/record.url?scp=0032208403&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0032208403&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:0032208403

VL - 12

SP - 129

EP - 153

JO - Machine Learning

JF - Machine Learning

SN - 0885-6125

IS - 4

ER -