• 论文 •    

最近邻实例检索相似度分析

蒋占四,陈立平,罗年猛   

  1. 华中科技大学 机械科学与工程学院CAD中心,湖北武汉430074
  • 出版日期:2007-06-15 发布日期:2007-06-25

Similarity analysis in nearestneighbor case retrieval

JIANG  Zhansi, CHEN Liping, LUO Nianmeng   

  1. CAD Center, School of Mechanical Science & Engineering, Huazhong University of Science & Technology, Wuhan430074, China
  • Online:2007-06-15 Published:2007-06-25

摘要: 针对最近邻实例检索中实例属性相似度和权重的计算问题,给出了区间值属性相似度的计算模型,并将各种属性类型的相似度计算方法加以统一,提出了基于相似度离差信息的客观赋权方法,并以组合权重计算实例相似度。该方法已应用于基于实例推理的导弹概念设计之中,结果表明,所提出的权重计算方法更有利于实例的检索。

关键词: 实例推理, 实例检索, 相似度

Abstract: To deal with problems of similarity measurement and weights assignment in CaseBased Reasoning (CBR), a similarity computation model between range properties was introduced. This model unified various properties computation methods. To assign weights, an objective method based on the deviation information of similarity values among properties was proposed. Objective and subjective weights were combined to form synthesis weights. Finally, the similarity measurement and weights assignment methods were applied in a missile conception design CBR system. It was demonstrated that this method could improve case retrieval.

Key words: casebased reasoning, case retrieval, similarity, weight

中图分类号: