三元规则强度推理的模糊数据模型

A REASONING MODEL WITH 3-FACTOR RULE STRENGTH

本文提出了一种不精确推理的数学模型,该模型将MYCIN推理的可信度和Prospector的主观Bayes推理模型的充分性和必要性因子相结合,形成了一种不必赋先验概率的具有三元规则强度的推理模型。在此基础上,给出一种模糊关系的数据模型,并定义了模糊集合的并、交、差和笛卡尔积等运算,从而对非精确数据及知识描述中的不确定性进行有效地处理。

This paper presents a mathematical model for inacuracy reasoning. This model combines the CF of MYCIN reasoning and LS and LN of Prospector's Bayes reasoning into a kind of reasoning that is not necessarily assigned by prior probabilities and has 3 - factor rule strength. Besides, this paper defines the operations of AND,OR,SUBSTRACT and D- Multiply on the fuzzy set. So one can suffienciently deal with the uncertainty in the description of inaccurate data and knowledge.