arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1905.05348cs.LGstat.ML
收藏

谱近似推理

Spectral Approximate Inference

KAIST(韩国科学技术院);AITRICS

Sejun Park, Eunho Yang, Se-Young Yun, Jinwoo Shin

摘要:

给定一个图模型(GM),计算其配分函数是最基本的推理任务,但它通常是计算上难以处理的。为了解决这个问题,迭代近似算法探索某些局部结构/一致性的GM已被调查,在实践中流行的选择。然而,由于它们的局部/迭代性质,它们通常输出差的近似值或甚至不收敛,例如,在低温状态下(大参数的硬实例)。为了克服这一局限性,我们提出了一种新的方法,利用GM的全局谱特征。我们的贡献是双重的:(a)首先提出了一种全多项式时间近似方案(FPTAS),用于近似与低秩耦合矩阵相关联的GM的配分函数;(b)对于一般的高阶GM,我们设计了一种谱平均场方案,利用(a)作为子程序,将高阶GM近似为秩1 GM的乘积,以有效地近似配分函数。所提出的算法在其运行时间和准确性方面比现有方法更鲁棒,即,既不受收敛问题的影响,也不依赖于硬局部结构,如我们的实验所示。

英文摘要:

Given a graphical model (GM), computing its partition function is the most essential inference task, but it is computationally intractable in general. To address the issue, iterative approximation algorithms exploring certain local structure/consistency of GM have been investigated as popular choices in practice. However, due to their local/iterative nature, they often output poor approximations or even do not converge, e.g., in low-temperature regimes (hard instances of large parameters). To overcome the limitation, we propose a novel approach utilizing the global spectral feature of GM. Our contribution is two-fold: (a) we first propose a fully polynomial-time approximation scheme (FPTAS) for approximating the partition function of GM associating with a low-rank coupling matrix; (b) for general high-rank GMs, we design a spectral mean-field scheme utilizing (a) as a subroutine, where it approximates a high-rank GM into a product of rank-1 GMs for an efficient approximation of the partition function. The proposed algorithm is more robust in its running time and accuracy than prior methods, i.e., neither suffers from the convergence issue nor depends on hard local structures, as demonstrated in our experiments.

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。

↑