和积网络中的最大后验推理
Maximum A Posteriori Inference in Sum-Product Networks
摘要:
和积网络(SPN)是一类允许易于处理的边际推理的概率图模型。然而,最大后验概率(MAP)推理在SPN是NP难的。我们从理论和算法的角度研究了SPN中的MAP推理。对于理论部分,我们减少一般MAP推理的特殊情况下,没有证据和隐藏变量,我们还表明,它是NP-困难的近似MAP问题$2^{n^ε}$固定$0 \leq ε< 1$,其中$n$是输入大小。对于算法部分,我们首先提出了一个精确的MAP求解器,运行速度相当快,可以处理SPN与多达1 k个变量和150 k弧在我们的实验。然后,我们提出了一种新的近似MAP求解器,在速度和精度之间具有良好的平衡,我们在真实世界数据集上的综合实验表明,它具有更好的整体性能比现有的近似求解器。
英文摘要:
Sum-product networks (SPNs) are a class of probabilistic graphical models that allow tractable marginal inference. However, the maximum a posteriori (MAP) inference in SPNs is NP-hard. We investigate MAP inference in SPNs from both theoretical and algorithmic perspectives. For the theoretical part, we reduce general MAP inference to its special case without evidence and hidden variables; we also show that it is NP-hard to approximate the MAP problem to $2^{n^ε}$ for fixed $0 \leq ε< 1$, where $n$ is the input size. For the algorithmic part, we first present an exact MAP solver that runs reasonably fast and could handle SPNs with up to 1k variables and 150k arcs in our experiments. We then present a new approximate MAP solver with a good balance between speed and accuracy, and our comprehensive experiments on real-world datasets show that it has better overall performance than existing approximate solvers.
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。