arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1202.3777cs.AIcs.DC
收藏

通过连接树中的消息传递的置信传播:使用GPU并行化更快地计算每个消息

Belief Propagation by Message Passing in Junction Trees: Computing Each Message Faster Using GPU Parallelization

Carnegie Mellon University(卡内基梅隆大学);University of California, Berkeley(加州大学伯克利分校)

Lu Zheng, Ole Mengshoel, Jike Chong

摘要:

将贝叶斯网络(BN)编译为连接树并在其上执行置信传播是计算BN中后验的最突出方法之一。然而,连接树上的信念传播是已知的,在一般情况下是计算密集型的。它的复杂性可能会显着增加贝叶斯网络节点的连通性和状态空间基数。在本文中,我们使用GPU并行化来解决这个计算挑战。我们开发的数据结构和算法,扩展现有的连接树技术,并专门开发一种新的方法来计算每个信念传播消息并行。我们实现了我们的方法上的NVIDIA GPU和测试它使用BN从几个应用程序。在实验中,我们研究了连接树参数如何影响并行化的机会,因此我们的算法的性能。我们实现了从0.68到9.18的BN研究的加速比。

英文摘要:

Compiling Bayesian networks (BNs) to junction trees and performing belief propagation over them is among the most prominent approaches to computing posteriors in BNs. However, belief propagation over junction tree is known to be computationally intensive in the general case. Its complexity may increase dramatically with the connectivity and state space cardinality of Bayesian network nodes. In this paper, we address this computational challenge using GPU parallelization. We develop data structures and algorithms that extend existing junction tree techniques, and specifically develop a novel approach to computing each belief propagation message in parallel. We implement our approach on an NVIDIA GPU and test it using BNs from several applications. Experimentally, we study how junction tree parameters affect parallelization opportunities and hence the performance of our algorithm. We achieve speedups ranging from 0.68 to 9.18 for the BNs studied.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑