摘要
分析了传统的实例检索策略的不足,提出了一种区间属性值之间相似度的计算方法.该方法是通过计算长度重叠率或面积重叠率来计算其相似程度.所提出的方法可以有效地处理实例检索中的不确定性,并能提高基于实例推理系统的性能.
The disadvantages of the traditional cases indexing models are analyzed. A new kind of method for range attributes similarity calculation is presented. This method uses Overlap rate of length or area to calculato the similarity between two range attributes. The method can not only process the imprecise information in the case retrieval, but also improve the performance of case-based reasoning system.
出处
《辽宁师范大学学报（自然科学版）》
CAS
北大核心
2006年第4期442-445,共4页
Journal of Liaoning Normal University:Natural Science Edition
基金
教育部科学基金资助项目(KP0302
2003-2005)
关键词
实例推理
相似度
确定区间
模糊区间
case-based reasoning
similarity
certain range
uncertain range
