GP-Select: Accelerating EM Using Adaptive Subspace Preselection

GP-Select: Accelerating EM Using Adaptive Subspace Preselection
复制标题

GP-Select:使用自适应子空间预选加速 EM

DOI:
10.1162/neco_a_00982
复制
发表时间:
2017
期刊:
影响因子:
2.9
通讯作者:
Arthur Gretton
Arthur Gretton
中科院分区:
计算机科学4区
文献类型:
--
作者:
Jacquelyn A Shelton;Jan Gasthaus;Zhenwen Dai;Jörg Lücke;Arthur Gretton

文献摘要

参考文献

被引文献

相似文献

我们提出了一个非参数的过程,以实现快速推理生成图形模型时,潜在状态的数量是非常大的。该方法是基于迭代潜变量预选,在这里我们交替学习选择函数,以揭示相关的潜变量,并使用它来获得一个紧凑的近似EM的后验分布。这可以使得推断成为可能,其中可能的潜在状态的数量例如是潜在变量的数量的指数,而精确的方法在计算上是不可行的。我们通过高斯过程回归完全从观测数据和当前期望最大化状态中学习选择函数。这与早期的方法形成对比,在早期的方法中,为每个问题设置手动设计选择函数。我们表明,我们的方法执行以及这些定制的选择功能上的各种各样的推理问题。特别是,对于具有挑战性的情况下,层次模型的对象定位与遮挡,我们实现了结果,匹配定制的最先进的选择方法,在一个低得多的计算成本。
We propose a nonparametric procedure to achieve fast inference in generative graphical models when the number of latent states is very large. The approach is based on iterative latent variable preselection, where we alternate between learning a selection function to reveal the relevant latent variables and using this to obtain a compact approximation of the posterior distribution for EM. This can make inference possible where the number of possible latent states is, for example, exponential in the number of latent variables, whereas an exact approach would be computationally infeasible. We learn the selection function entirely from the observed data and current expectation-maximization state via gaussian process regression. This is in contrast to earlier approaches, where selection functions were manually designed for each problem setting. We show that our approach performs as well as these bespoke selection functions on a wide variety of inference problems. In particular, for the challenging case of a hierarchical model for object localization with occlusion, we achieve results that match a customized state-of-the-art selection method at a far lower computational cost.
DOI: 10.5555/2627435.2697053
发表时间: 2014
期刊: J. Mach. Learn. Res.
影响因子: --
作者:
M. Henniges;Richard E. Turner;M. Sahani;J. Eggert;Jörg Lücke
通讯作者: M. Henniges;Richard E. Turner;M. Sahani;J. Eggert;Jörg Lücke
GPS-ABC:高斯过程代理近似贝叶斯计算
DOI: --
发表时间: 2014
期刊: Conference on Uncertainty in Artificial Intelligence
影响因子: --
作者:
Edward Meeds;M. Welling
通讯作者: M. Welling
DOI: 10.1111/j.2517-6161.1977.tb01600.x
发表时间: 1977-01-01
期刊: JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES B-METHODOLOGICAL
影响因子: --
作者:
DEMPSTER, AP;LAIRD, NM;RUBIN, DB
通讯作者: RUBIN, DB
DOI: --
发表时间: 1999
期刊: Neural Networks
影响因子: 7.8
作者:
E. Körner;M. Gewaltig;Ursula Körner;A. Richter;Tobias Rodemann
通讯作者: Tobias Rodemann
DOI: 10.1371/journal.pone.0124088
发表时间: 2015
期刊: PloS one
影响因子: 3.7
作者:
Shelton JA;Sheikh AS;Bornschein J;Sterne P;Lücke J
通讯作者: Lücke J