A improved CF model base on normal distribution and Euclidean distance

Abstract

Since the CF model has been proposed, it has been successfully applied in some areas and there is one of the representative systems MYCIN. However, the traditional CF model also has a few problems, for example, the inconsistencies of the degree of belief in a hypothesis and Conditional probability to some extent. Then this paper will proposed an improved method of the CF model which is the improved CF model base on normal distribution and Euclidean distance. It can resolve the serious inconsistencies of the degree of belief in a hypothesis and Conditional probability to some extent effectively.