计算机应用 ›› 2012, Vol. 32 ›› Issue (04): 1030-1032.DOI: 10.3724/SP.J.1087.2012.01030

• 人工智能 • 上一篇    下一篇

基于蚁群聚类的历史灾害分级方法

贾志娟1,胡明生2,刘思2   

  1. 1. 武汉理工大学 计算机科学与技术学院,武汉 430070
    2. 郑州师范学院 网络中心,郑州 450044
  • 收稿日期:2011-10-17 修回日期:2011-12-07 发布日期:2012-04-20 出版日期:2012-04-01
  • 通讯作者: 刘思
  • 作者简介:贾志娟(1973-),女,河南郑州人,博士研究生,主要研究方向:人工智能、数据挖掘;胡明生(1973-),男,河南新县人,副教授,博士,主要研究方向:智能控制与决策支持;刘思(1985-),男,河南信阳人,硕士研究生,主要研究方向:人工智能。
  • 基金资助:
    国家自然科学基金资助项目;河南省基础与前沿技术研究项目;河南省教育厅自然科学研究计划项目

A Historical Disaster Classification Method Based on Ant Colony Clustering

JIA Zhi-juan1,HU Ming-sheng2,LIU Si2   

  1. 1. School of Computer Science and Technology, Wuhan University of Technology, Wuhan Hubei 430070, China
    2. Network Center, Zhengzhou Normal University, Zhengzhou Henan 450044, China
  • Received:2011-10-17 Revised:2011-12-07 Online:2012-04-20 Published:2012-04-01
  • Contact: LIU Si
  • Supported by:
    ;the Natural Science Research Projects of Education Department of Henan Province

摘要: 针对历史灾害记录的描述性、简约性问题,提出一种基于蚁群聚类的历史灾害分级方法。利用灰色关联分析方法对灾害数据进行归一化处理后,再通过蚁群自动聚类的结果来划分历史灾害的等级,以避免人为的主观任意性干扰。通过与其他分级方法的性能对比,实验结果证明该方法具有较高的精确性和实用性。

关键词: 历史灾害, 分级, 聚类, 蚁群算法, 灰色关联分析

Abstract: Aiming at the descriptiveness and parsimony problems of historical disaster records a historical disaster classification method which based on ant colony clustering is proposed in this paper. The disaster data are normalized by using gray relational analysis approach, and then the levels of historical disasters are divided by the results of ant colony automatic clustering, so as to avoid the arbitrary man-made interference. In comparison with other classification methods in performance, experimental results show that this method has high accuracy and practicality.

Key words: historical disasters, classification, clustering, ant colony algorithm, gray relational analysis