Certainty factor theory: Its probabilistic interpretations and problems

摘要

In this article, some topics related with the MYCIN model are discussed. It is pointed out that some problems identified by Adams should be considered as evidence supporting the following point of view: the probabilistic interpretation for CFs given by Shortliffe is inappropriate. In connection to this, some results achieved by Heckerman are examined. The correctness of one of his arguments is questioned, and it has been found out that problems similar to those identified by Adams exist also in the interpretation given by Heckerman. Then, in the context of the result achieved by Horvitz and Heckerman, it is pointed out that the evidence combination scheme for the likelihood ratio suffers from another restriction. The reasonableness of an assumption made in PROSPECTOR is also questioned. At last, some short remarks are made with respect to the complaint that the MYCIN model does not work well in some other systems.