摘要
用传统的最近邻检索和匹配算法找到的最相似范例,由于时间的变迁,在实际应用中,该范例的解决方案对于 新问题的解决不一定是最有用的,因而也不是最优的.为了解决这个问题,本文提出了一种基于时序的范例匹配方法模 型,为找到最优相似范例提供了一条有效途径.
The similarist case which has been found in the existing nearest-neighbor retrieving and matching algorithms is probably useless to solve new problem. It is possibly because in real applications, situation changes as time goes on. To solve the problem,the paper proposes a new model of case matching method based on time-serial in case-basesd reasoning and provides an efficient way of finding the best similar case.
出处
《小型微型计算机系统》
CSCD
北大核心
2003年第4期640-642,共3页
Journal of Chinese Computer Systems
基金
国家自然科学基金项目(No.60075015)资助
