基于案例属性特征的相似度计算模型研究  

Research of the Similarity Calculation Models Based on the Features of Case Properties

作　　者：周凯波[1] 冯珊[1] 李锋[1] 

机构地区：[1]华中科技大学系统工程研究所,湖北武汉430074

出　　处：《武汉理工大学学报（信息与管理工程版）》2003年第1期24-27,共4页Journal of Wuhan University of Technology：Information & Management Engineering

基　　金：国家自然科学基金"九五"重大资助项目(79990 5 80 )

摘　　要：传统案例属性之间的相似度计算模型缺乏对不确定环境中不精确信息的适应性。引入了区间数的概念来描述案例的不精确属性 ,同时基于在某一区间之内的两精确属性间的相似度计算模型 ,提出了一种包含不精确信息的案例属性之间的相似度计算模型。所提出的模型可以有效地处理案例检索中的不确定性 。Traditional models, which calculate similarities between case properties, lack of the adaptation to imprecise information in an uncertain environment. Ranges are introduced to describe imprecise properties. Models, which can calculate similarities between case properties including imprecise information based on the similarities calculation models between two precise properties in a certain range, are presented. These models not only can process the imprecise information in the case retrieval, but also can improve the performance of the case-based reasoning system.

关 键 词：相似度计算模型 案例属性 案例推理 案例检索 不精确属性 特征属性 区间属性 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]