论文部分内容阅读
针对传统关联规则和 Boolean规则矩阵无法进行不确定性知识表示和推理的弱点,该文采用三值逻辑表示规则和真值状态的不确定性,提出并使用三值关联规则(T规则)和三值规则矩阵(TR矩阵),利用 TR矩阵变换实现高速前向推理和后向推理。对复合三值规则的无损分解进行了详细探讨,最后提出了一个能进行不确定推理的新算法。
Aiming at the weakness that the traditional association rules and the Boolean rule matrix can not carry on the indefinite knowledge representation and reasoning, this paper uses the three-valued logic to express the uncertainty of the rule and truth state, and proposes and uses three-valued association rules (T rules) Value rule matrix (TR matrix), using TR matrix transformation to achieve high-speed forward reasoning and backward reasoning. In this paper, the non-destructive decomposition of compound three-value rule is discussed in detail. Finally, a new algorithm that can make uncertain reasoning is proposed.