arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1401.1247cs.AI
收藏

通过可交换性实现可追踪性:有效概率推理的新视角

Tractability through Exchangeability: A New Perspective on Efficient Probabilistic Inference

University of Washington(华盛顿大学);University of California, Los Angeles(加州大学洛杉矶分校)

Mathias Niepert, Guy Van den Broeck

摘要:

互换性是统计学和概率论中的一个中心概念。无限序列的数据点是可交换的假设是贝叶斯统计的核心。然而,有限交换作为一个统计属性,使概率推理听话是不太好理解。我们发展了一个有限交换理论及其与易处理的概率推理的关系。该理论是对独立性和条件独立性理论的补充。我们表明,易于处理的推理概率模型与高树宽和数以百万计的变量可以理解使用有限(部分)交换的概念。我们还表明,现有的提升推理算法隐含地利用条件独立性和部分交换的组合。

英文摘要:

Exchangeability is a central notion in statistics and probability theory. The assumption that an infinite sequence of data points is exchangeable is at the core of Bayesian statistics. However, finite exchangeability as a statistical property that renders probabilistic inference tractable is less well-understood. We develop a theory of finite exchangeability and its relation to tractable probabilistic inference. The theory is complementary to that of independence and conditional independence. We show that tractable inference in probabilistic models with high treewidth and millions of variables can be understood using the notion of finite (partial) exchangeability. We also show that existing lifted inference algorithms implicitly utilize a combination of conditional independence and partial exchangeability.

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。

↑