Skip to main content

Research Repository

Advanced Search

Escaping local optima in multi-agent oriented constraint satisfaction.

Basharu, Muhammed; Ahriz, Hatem; Arana, In�s

Authors

Muhammed Basharu



Contributors

Frans Coenen
Editor

Alun Preece
Editor

Ann Macintosh
Editor

Abstract

We present a multi-agent approach to constraint satisfaction where feedback and reinforcement are used in order to avoid local optima and, consequently, to improve the overall solution. Our approach, FeReRA, is based on the fact that an agents local best performance does not necessarily contribute to the systems best performance. Thus, agents may be rewarded for improving the systems performance and penalised for not contributing towards a better solution. Hence, agents may be forced to choose sub-optimal moves when they reach a specified penalty threshold as a consequence of their lack of contribution towards a better overall solution. This may allow other agents to choose better moves and, therefore, to improve the overall performance of the system. FeReRA is tested against its predecessor, ERA, and a comparative evaluation of both approaches is presented.

Citation

BASHARU, M., AHRIZ, H. and ARANA, I. 2004. Escaping local optima in multi-agent oriented constraint satisfaction. In Coenen, F., Preece, A. and Macintosh, A. (eds.) Research and development in intelligent systems XX: technical proceedings of the 23rd Annual international conference of the British Computer Society's Specialist Group on Artificial Intelligence (SGAI) (AI-2003), 15-17 December 2003, Cambridge, UK. London: Springer [online], pages 97-110. Available from: https://doi.org/10.1007/978-0-85729-412-8_8

Conference Name 23rd Annual international conference of the British Computer Society's Specialist Group on Artificial Intelligence (SGAI) (AI-2003)
Conference Location Cambridge, UK
Start Date Dec 15, 2003
End Date Dec 17, 2003
Acceptance Date Jul 31, 2003
Online Publication Date Dec 31, 2004
Publication Date Dec 31, 2004
Deposit Date Jan 27, 2009
Publicly Available Date Jan 27, 2009
Publisher Springer
Pages 97-110
ISBN 9781852337803
DOI https://doi.org/10.1007/978-0-85729-412-8_8
Public URL http://hdl.handle.net/10059/298

Files





You might also like



Downloadable Citations