| 摘要: |
| 为使概率逻辑的不确定性推理方法能应用于较大规模的知识库，本文基于一个实际专家系统知识库的开发经验，在概率逻辑公式一致性区间的一般算法基础上，为概率逻辑公式集的分解设计了一种合并聚类算法．对于不同背景的概率逻辑知识库，只要公式集具有一定的分层结构性质，该算法就能保证Ｄａｎｔｚｉｇ－Ｗｏｌｆｅ分解的联合计算模型适用于概率逻辑推理．测试结果表明，该算法对于数１０个变量和子句的实例可收到很好的效果． |
| 关键词: 概率逻辑 一致性 分解算法 聚类 专家系统 概率推理 |
| DOI： |
| 分类号: |
| 基金项目:本文研究得到国家自然科学基金、国家863高科技项目资助. |
|
| THE REPRESENTATION AND CLUSTERING OF NUMERIC ATTRIBUTES IN CONCEPT FORMATION |
|
ZHANG Chendong,CHEN Huowang,LIU Fenqi
|
| Abstract: |
| In order that the probabilistic logic reasoning under uncertainty can be used for 1arge scale knowledge-base，this paper presents a clustering-algorithm for decomposi-tion of probabilistic logic formula set based on the general consistence assigning algorithm for Drobabilistic logic and the experience of developing the knowledge-base in an practical expert system． It ensures that the united-model of Dantzig-Wolfe decomposition can be used for probabilistic logic reasoning on probabilistic logic knowledge-base with different background，provided the formula set hold certain hierarchical structure．Experiments show that the algorithm performs successfully on instances with dozens of variables and clauses． |
| Key words: Probabilistic logic consistency decomposition algorithm clustering expert system probabilistic reasoning |
