Community-based multi-agent cooperative interactive evolutionary computation model. (1st January 2013)

• Record Type: Journal Article
• Authors: Guo, Yinan; Cheng, Jian; Chen, Meirong; Liu, Dandan
• Is Part Of: International journal of information and communication technology. Volume 5:Number 3/4(2013)
• Page Start: 243
• Page End: 256
• Publication Date: 2013-01-01
• Subjects: community -- cooperative -- interactive evolutionary algorithms -- IEAs -- agent; Information technology -- Periodicals; Computer science -- Periodicals; Telecommunication -- Periodicals; 004.05
• Languages: English
• ISSNs: 1466-6642

Abstract

Traditional cooperative interactive evolutionary algorithms are usually implemented based on local area network or parallel computer system. It cannot fully utilise the resources in wide area network. To solve this problem, a community-based multi-agent cooperative interactive evolutionary computation model is constructed based on the autonomous agents. Users having similar preferences are partitioned into one community, whereas they participate in the algorithm at the different computation nodes of wide area network. In a community, users communicated according to their needs. Knowledge among communities is shared by bulletin board. Through this hierarchical structure for information exchange, each user can fully utilise other users' evolution information to speed-up the evolution itself. Moreover, it can effectively reduce communication costs and avoid network congestion. Taking fashion evolutionary design system as an example, experimental results indicate that the structure of proposed model is rational. It has less communication cost and can greatly alleviate human fatigue.