一种改进的CBR案例检索相似性度量模型

案例相似性度量是基于案例推理（CBR）的突发事件应急决策管理的关键问题。本文基于最近邻算法,提出一种改进的案例相似性度量模型。提出＂结构相似度＂度量缺失数据对于案例结构特征相似性的影响,引入＂变异系数＂度量案例属性的可替代性,精确描述替代程度不同的属性对于案例相似性的贡献。最后通过具体案例检索实例验证了该模型的有效性。

An Improved Case Similarity Measure for Case Retrieval in CBR Based on Emergency Decision

Case-Based Reasoning（CBR） is widely used in emergency decision.Case similarity measure is a key issue in the case retrieval process.In this paper,an improved case similarity measure is given to deal with the problem of missing data and substitutability,which are ignored by traditional case similarity measures.＇Structure similarity＇ is proposed to calculate the structural impact of missing data,and ＇variation coefficient＇ is used to measure the substitutability between attributes.Finally,numerical examples in the real emergency decision application justified the effectiveness of the new case similarity measure.