基于强相关逻辑的贝叶斯网络结构学习方法  

Structure Learning of Bayesian Networks Based on Strong Relevant Logics

作　　者：赵建喆[1] 李凯[1] 

机构地区：[1]东北大学工商管理学院,沈阳110819

出　　处：《情报学报》2014年第11期1204-1214,共11页Journal of the China Society for Scientific and Technical Information

基　　金：1)基金项目国家自然科学基金(61374178)。

摘　　要：对问题域的经验知识进行知识发现并形成贝叶斯网络结构是贝叶斯网络建模的核心问题。但是现有的结构学习方法的底层逻辑系统不具备知识发现的能力。强相关逻辑（strong relevant logic，SRL）系统的相关推论是获取新知识的有效手段。为了解决传统贝叶斯网络结构学习方法的主观性问题，本文提出一种新的贝叶斯网络结构学习方法。该方法首先建立了一种强相关逻辑为基础的贝叶斯网络形式化的表达系统，给出了方法实现的流程及其子算法，最后通过实例分析该方法进行结构建模的有效性。The key issue of Bayesian network modeling is discovering empirical knowledge from problem domain, based on which Bayesian network is constructed. However, the basis logical system of existing structure learning methods cannot discover knowledge. Moreover, relevant inference of strong relevant logic （SRL） is an effect method for obtaining new knowledge. This paper proposes a new method for Bayesian networks structure learning, which solves the subjective problem of traditional learning methods. Expression system that formalized by Bayesian networks is built based on SRL. The processing method and its inner Mgorithms are also described in this paper. The results of cases studying demonstrate the effectiveness of the proposed methods.

关 键 词：贝叶斯网络 强相关逻辑 结构学习 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]