arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1301.7406cs.AI
收藏

对数时间并行贝叶斯推理

Logarithmic Time Parallel Bayesian Inference

University of Michigan AI Laboratory(密歇根大学人工智能实验室)

David M. Pennock

摘要:

本文提出了一种贝叶斯网络精确概率推理的并行算法。对于具有n个变量的多叉树网络,在具有n个处理器的CREW PRAM(并发读、异写并行随机存取机)上,对于任何常数数量的证据变量,最坏情况下的时间复杂度为O(log n)。对于任意网络,时间复杂度对于n个处理器是O(r ^{3w}* log n),或者对于r ^{3w}* n个处理器是O(w * log n),其中r是任何变量的最大范围,w是在道德化和三角化网络之后的诱导宽度(最大团大小)。

英文摘要:

I present a parallel algorithm for exact probabilistic inference in Bayesian networks. For polytree networks with n variables, the worst-case time complexity is O(log n) on a CREW PRAM (concurrent-read, exclusive-write parallel random-access machine) with n processors, for any constant number of evidence variables. For arbitrary networks, the time complexity is O(r^{3w}*log n) for n processors, or O(w*log n) for r^{3w}*n processors, where r is the maximum range of any variable, and w is the induced width (the maximum clique size), after moralizing and triangulating the network.

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。

↑