• 论文 •    

基于贝叶斯网络的不确定性知识的推理方法

胡玉胜,涂序彦,崔晓瑜,程乾生   

  1. 1.北京科技大学信息工程学院自动化系,北京100083;2.北京大学数学科学学院信息科学系,北京100871
  • 出版日期:2001-12-15 发布日期:2001-12-25

A Inferential Method of Uncertain Knowledge Based on Bayes-Network

HU Yu-sheng,TU Xu-yan,CUI Xiao-yu,CHENG Qian-sheng   

  1. 1. Information Engineering Institute,Beijing University of Science and Technology,Beijing100083,China;2.Mathematical Science Institute,Beijing University,Beijing100871,China
  • Online:2001-12-15 Published:2001-12-25

摘要: 贝叶斯网络是不确定性知识表达与推理的一种新方法。它是概率论和图论相结合的产物,可用于复杂多因果关系的分析,是人工智能领域的研究热点和重要成果之一。由于它的解决方案明确、直观,所以近年来在远程医疗、故障诊断以及数据挖掘等领域,得到了广泛的应用。本文论述了贝叶斯网络的基本理论、方法和应用,并指出当前所存在的主要问题。

关键词: 贝叶斯网络, 人工智能, 条件无关性, 因果关系, 因果推理, 诊断推理

Abstract: Bayes Network is a new inference and express method of uncertain knowledge.It is the combination of probability and graph theory, and can be used in analyzing complicated multi-causality questions. It is one of studying hotspots and important fruits in artificial intelligence. Because of its explicit and intuitionist, it has been widely used in many fields, such as internal medicine, diagnosing, and data mining. This paper addressed Bayes network's basic theory, methods and applications, finally pointed out its major problems at present.

Key words: bayes network, artificial intelligence, conditional independence, causality, causal inference, diagnostic inference

中图分类号: