Parallel and distributed multi-agent reinforcement learning

Author

Kaya, Mehmet ; Arslan, Ahmet

Author_Institution

Dept. of Comput. Eng., Firat Univ., Elazig, Turkey

fYear

2001

fDate

2001

Firstpage

437

Lastpage

441

Abstract

The application of parallel and distributed systems to multi-agent environments has attracted recent attention. Multi-agent systems are a particular type of distributed artificial intelligence system. This paper presents an approach to learning in parallel and distributed systems. A variant of the job assignment problem is chosen as an evaluation task. This is an NP-hard problem, which is relevant to many industrial application domains. Experimental results show the effectiveness of the proposed approach

Keywords

computational complexity; distributed processing; learning (artificial intelligence); multi-agent systems; production control; scheduling; NP-hard problem; distributed artificial intelligence; distributed systems; industrial applications; job assignment problem; multi-agent reinforcement learning; parallel systems; Application software; Artificial intelligence; Concurrent computing; Control systems; Distributed computing; Intelligent robots; Learning; Multiagent systems; Parallel processing; Robustness;

fLanguage

English

Publisher

ieee

Conference_Titel

Parallel and Distributed Systems, 2001. ICPADS 2001. Proceedings. Eighth International Conference on

Conference_Location

Kyongju City

ISSN

1521-9097

Print_ISBN

0-7695-1153-8

Type

conf

DOI

10.1109/ICPADS.2001.934851

Filename

934851