arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1303.1499cs.AI
收藏

用树分解结构逼近信度网络

Using Tree-Decomposable Structures to Approximate Belief Networks

Louisiana State University(路易斯安那州立大学)

Sumit Sarkar

摘要:

树结构已被证明为传播信念提供了一个有效的框架[Pearl,1986]。本文研究了寻找最优逼近树的问题。三个二元变量的星分解方案[Lazarsfeld,1966; Pearl,1986]被证明可以增强支持树结构的概率分布类,这种结构被称为树可分解结构。对数评分规则被发现是一个适当的最优性标准,以评估不同的树可分解的结构。最接近于实际的信念网络的结构的特性,确定使用对数规则,贪婪和精确的技术来找到最佳的近似。

英文摘要:

Tree structures have been shown to provide an efficient framework for propagating beliefs [Pearl,1986]. This paper studies the problem of finding an optimal approximating tree. The star decomposition scheme for sets of three binary variables [Lazarsfeld,1966; Pearl,1986] is shown to enhance the class of probability distributions that can support tree structures; such structures are called tree-decomposable structures. The logarithm scoring rule is found to be an appropriate optimality criterion to evaluate different tree-decomposable structures. Characteristics of such structures closest to the actual belief network are identified using the logarithm rule, and greedy and exact techniques are developed to find the optimal approximation.

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。

↑