计算机应用 ›› 2012, Vol. 32 ›› Issue (12): 3278-3282.DOI: 10.3724/SP.J.1087.2012.03278

• 先进计算 • 上一篇    下一篇

改进图聚类算法及其应用

丁利,向来生,刘希玉,宋超超   

  1. 山东师范大学 管理科学与工程学院, 济南 250014
  • 收稿日期:2012-06-27 修回日期:2012-08-06 发布日期:2012-12-29 出版日期:2012-12-01
  • 通讯作者: 丁利
  • 作者简介:丁利(1988-),女,山东禹城人,硕士,主要研究方向:数据挖掘;〓向来生(1957-),男,山东淄博人,研究员,主要研究方向:可持续发展;〓刘希玉(1964-),男,山东莱芜人,教授,博士生导师,博士,主要研究方向:电子商务、计算智能;〓宋超超(1988-),男,山东济宁人,硕士,主要研究方向:计算智能。
  • 基金资助:
    国家自然科学基金资助项目;山东省自然科学基金项目

Improved graph clustering algorithm and its application

Li Ding,XIANG Lai-sheng,LIU Xi-yu,SONG Chao-chao   

  1. Institute of Management Science and Engineering, Shandong Normal University,Jinan Shandong 250014, China
  • Received:2012-06-27 Revised:2012-08-06 Online:2012-12-29 Published:2012-12-01
  • Contact: Li Ding

摘要: 第四方物流企业联盟建立问题是研究如何将区域内物流企业以一种高效、低联系代价的方式建立合作联盟的问题。针对该问题提出一种基于离散粒子群优化算法的改进图聚类算法,有助于降低合作联盟之间的联系代价。通过离散粒子群算法优化基本图聚类算法得到的初期聚类结果,利用扰动策略对优化结果进行再拓展。对于实验中的100家虚拟企业进行了聚类,使得联盟企业内部总联系代价从初始时的39991降低到最后的24800。实验结果表明,基于离散粒子群算法的改进图聚类算法能以较低的花费解决物流企业联盟建立问题。

关键词: 第四方物流, 企业联盟, 图聚类, 粒子群算法, 离散化, 扰动策略

Abstract: The fourth-party logistics company alliance building problem is to study how to build alliance with a method of high efficiency and low cost. An improved algorithm about graph clustering based on Particle Swarm Optimization (PSO) was proposed to solve the problem, and it contributed to reduce the connection cost between companies in the group. Discretization PSO was used to optimize the clustering result of graph clustering algorithm, and the result with disturbance strategies got enhanced. Clustering the 100 virtual companies, the cost was reduced from 39991 to 24800. The experimental result shows that, the improved graph clustering algorithm based on discretization PSO can solve the problem in a way with high efficiency and low cost.

Key words: the fourth party logistics, company alliance, graph clustering, PSO, discretization, disturbance strategies