[1]肖旭,慕德俊,张慧翔,等. GPU加速的贝叶斯网络精确推理方法研究[J].计算机技术与发展,2014,24(10):1-5.
　XIAO Xu,MU De-jun,ZHANG Hui-xiang,et al. Research on Bayesian Network’s Exact Inference Using GPU Acceleration[J].,2014,24(10):1-5.
点击复制

 GPU加速的贝叶斯网络精确推理方法研究()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
24
期数:
2014年10期
页码:
1-5
栏目:
智能、算法、系统工程
出版日期:
2014-10-10

文章信息/Info

Title:
 Research on Bayesian Network’s Exact Inference Using GPU Acceleration
文章编号:
1673-629X（2014）10-0001-05
作者:
 肖旭; 慕德俊; 张慧翔; 陈春雷
 西北工业大学 自动化学院
Author(s):
 XIAO Xu; MU De-jun; ZHANG Hui-xiang; CHEN Chun-lei
关键词:
 ; 贝叶斯网络; 团树传播算法; GPU加速; 并行化信念传播
Keywords:
 ; Bayesian network; junction tree propagation algorithm; GPU acceleration; parallel belief propagation
分类号:
TP391
文献标志码:
A
摘要:
 对于复杂输入的贝叶斯网络，精确推理时间较长。文中针对贝叶斯网络精确推理中的团树传播算法，提出了一种基于CPU-GPU异构计算平台的并行化方法。首先研究团节点间信念势更新方式，提出了节点级并行化方法加速更新过程；其次，提出了利用计算复杂度的优先级队列方法，通过拓扑级并行化加速全局推理过程；最后，通过输入不同团树结构-线性结构、两分支二叉树结构和完全二叉树结构验证算法加速效果。实验结果表明，节点级并行化方法对线性结构有明显加速效果，拓扑级并行化对两分支二叉树和满二叉树结构有明显加速效果。
Abstract:
 For Bayesian network with complex input,the inferring is time-consuming. Aiming at the junction tree propagation algorithm of Bayesian network’ exact inference,a parallel method is proposed based on the CPU-GPU heterogeneous computing platform. Firstly, the updating method between junction tree nodes is investigated,and node-level parallelism method is proposed to accelerate the upda-ting. Secondly,the priority queue method is presented according to the computational complexity to achieve topological-level parallelism to accelerate the global inference process. Finally,speedup of the proposed method is verified on various input junction trees including lin-ear structure,two branches of binary tree structure and complete binary tree structure. The experimental results indicate that the node-level parallelism method can significantly accelerate the linear structure,and topological-level parallelism is effective for the two branches of binary tree and complete binary tree.

相似文献/References:

[1]宫义山 高媛媛.基于信息融合的诊断贝叶斯网络研究[J].计算机技术与发展,2009,(06):106.
　GONG Yi-shan,GAO Yuan-yuan.Diagnostic Bayesian Networks Research Based on Information Fusion[J].,2009,(10):106.
[2]奚海荣 马文丽 梁斌.基于贝叶斯网络SP算法的改进研究[J].计算机技术与发展,2009,(03):155.
　XI Hai-rong,MA Wen-li,LIANG Bin.Improvement of SP Algorithm Based on Bayesian Networks[J].,2009,(10):155.
[3]王峻 周孟然.一种基于MDL度量的选择性扩展贝叶斯分类器[J].计算机技术与发展,2007,(07):35.
　WANG Jun,ZHOU Meng-ran.A Selective Augmented Naive Bayesian Classifier Based on MDL Score[J].,2007,(10):35.
[4]程泽凯.基于TAN结构的启发式贝叶斯网络结构学习算法[J].计算机技术与发展,2007,(08):61.
　CHENG Ze-kai.BN Structure Learning Heuristic Algorithm Based on TAN Structure[J].,2007,(10):61.
[5]蒋望东 林士敏 鲁明羽[].基于BIC测度和混合遗传算法的BNC结构学习[J].计算机技术与发展,2007,(03):84.
　JIANG Wang-dong,LIN Shi-min,LU Ming-yu.Structure Learning of BNC Based on BIC and Hybrid Genetic Algorithms[J].,2007,(10):84.
[6]孙笑微 赵大宇 李晓毅 唐恒永.用于数据挖掘的TAN分类器的研究与应用[J].计算机技术与发展,2006,(11):140.
　suN Xiao-wei,ZHAO Da-yu,LI Xiao-yi,et al.Study and Application of TAN Classifier for Data Mining[J].,2006,(10):140.
[7]钟仕群 朱程荣 熊齐邦.一种基于贝叶斯网络的集成的故障定位模型[J].计算机技术与发展,2006,(12):13.
　ZHONG Shi-qun,ZHU Cheng-rong,XIONG Qi-bang.An Integrated Fault Localization Model Based on Bayesian Networks[J].,2006,(10):13.
[8]潘晓东[] 王春江 黄镇鸿 张国辉[].作战重心建模中的条件概率生成方法研究[J].计算机技术与发展,2011,(04):56.
　PAN Xiao-dong,WANG Chun-jiang,HUANG Zhen-hong,et al.Research on Method of Generating Conditional Probabilities in COG Modeling[J].,2011,(10):56.
[9]洪海燕.基于贝叶斯分类器的简历筛选模型[J].计算机技术与发展,2012,(07):85.
　HONG Hai-yan.Screening of Resumes Based on Bayesian Classifier[J].,2012,(10):85.
[10]吴绍兵.基于贝叶斯网络的信息提取技术研究[J].计算机技术与发展,2012,(11):225.
　WU Shao-bing.Research of Information Extraction Technique Based on Bayesian Network[J].,2012,(10):225.
[11]崔鹏帅,朱浩洋,任浩. 基于分层模型的组合服务故障定位算法[J].计算机技术与发展,2014,24(09):6.
　CUI Peng-shuai,ZHU Hao-yang,REN Hao. A Fault Location Algorithm of Composition Service Based on Hierarchical Model[J].,2014,24(10):6.
[12]李正浩,刘学军. 一种基于贝叶斯网络的作战重心评估模型[J].计算机技术与发展,2014,24(09):50.
　LI Zheng-hao,LIU Xue-jun. An Evaluation Model of Center of Gravity Based on Bayesian Network[J].,2014,24(10):50.
[13]王巍. 基于云参数贝叶斯网络的威胁评估方法[J].计算机技术与发展,2016,26(06):106.
　WANG Wei. An Threat Assessment Method Based on Cloud Parameters Bayesian Network[J].,2016,26(10):106.
[14]刘栋,宁玉富. 事件关联在证据链构造中的研究[J].计算机技术与发展,2016,26(12):107.
　LIU Dong,NING Yu-fu. Research on Event Correlation in Construction of Evidence Chain[J].,2016,26(10):107.
[15]林青[],戴慧珺[],任德旺[]. 基于贝叶斯网络的量化信任评估方法[J].计算机技术与发展,2016,26(12):132.
　LIN Qing[],DAI Hui-jun[],REN De-wang[]. A Quantitative Trust Assessment Method Based on Bayesian Network[J].,2016,26(10):132.
[16]高晓利[],李捷[][]. 基于模糊变结构动态贝叶斯网的目标识别方法[J].计算机技术与发展,2017,27(09):17.
　GAO Xiao-li[],LI Jie[][]. A Target Identification Method of Dynamic Bayesian Network with Fuzzy Variable Structure[J].,2017,27(10):17.

更新日期/Last Update: 2015-04-02