一种改进的多模块贝叶斯网络局部推理算法.pdfVIP

第36卷第9期 东 北 大 学 学 报 （ 自 然 科 学 版 ） Vol.36，No.9 2015 年 9 月 Journal ofNortheastern University（Natural Science） Sep.20 1 5 doi:10.3969/j.issn.1005 -3026.2015.09.008 一种改进的多模块贝叶斯网络局部推理算法 赵建喆，李 凯 ( 北大学 工商管理学院,辽宁沈阳 110819) 摘 ：针对多模块贝叶斯网络的局部推理的时间和空间复杂度高的问题,提出了一种改进的多模块贝 叶斯网络局部推理算法.该算法用面向对象语言重新定义了多模块贝叶斯网络模型,在联合树推理算法的基 础上结合图论中“顶点度”的概念对局部推理算法进行了优化,针对三角化结果不唯一的问题,给出了一种一 般性的解决方案,使三角化后的结果能够将消息传递得更快,有效地缩短推理时间.给出了算法的仿真实例并 进行实验分析,结果表明改进后的推理算法有效减小时间、空间复杂度. 关 键 词：多模块贝叶斯网络(MSBN);局部推理;联合树算法;顶点度;三角化 中图分类号：TP 18 文献标志码：A 文章编号：1005 -3026(2015)09 -1251 -05 An ImproVed Local Inference Algorithm for Multiply Sectioned Bayesian Networks ZHAO Jian-zhe ,LI Kai (School ofBusiness Administration,Northeastern University,Shenyang 110819, hina. orresponding author: ZHAO Jian-zhe,E-mail:zhaojz@swc.neu.edu.cn) Abstract:Due to the temporal and spatial complexity inthe local inference ofmultiply sectioned Bayesian networks (MSBN),an improved algorithm for the local inference of MSBN was proposed.The algorithm redefined the model of MSBN with an object-oriented language. ombined with the concept of vertex degree in graph theory,the algorithm was optimized based on thejoint tree algorithm. onsidering that the outcome of triangulation was not single,the improved algorithm offereda general solution,which helped to convey messagefasterand greatly shorten inference time.Finally,aninstance ofthealgorithmwas given forexperimentalanalysis, whose results showed that the improved inference algorithm significantly reduces both temporal and spatial complexity. Key words:multiply sectioned Bayesian network;local inference;joint tree algorithm;vertex degree;triangulation 20 世纪90 年代以来,贝叶斯网络(Bayesian MSBNs是贝叶斯网络的一种拓展,将大型的复杂 networks,BNs)成为人工智能、决策科学、