arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1704.01445stat.MLcs.NAmath.NAstat.CO
收藏

对数行列式的概率推断

Bayesian Inference of Log Determinants

University of Oxford(牛津大学);EURECOM(欧洲通信管理研究所)

Jack Fitzsimons, Kurt Cutajar, Michael Osborne, Stephen Roberts, Maurizio Filippone

AI总结:

本文提出将计算对数行列式视为贝叶斯推断问题,结合矩阵理论的界和随机迹估计的证据,以概率方式估计对数行列式及其不确定性,方法在性能上与现有方法相当并量化了预算受限的不确定性。

摘要:

核矩阵的对数行列式出现在多种机器学习问题中,从行列式点过程和广义马尔可夫随机场到高斯过程的训练。当核矩阵大小超过几千时,精确计算这一项通常是不可行的。在概率数值学的框架下,我们将计算对数行列式的问题重新解释为贝叶斯推断问题。具体来说,我们结合矩阵理论中的界和随机迹估计得到的证据,以在给定计算预算内获得对数行列式及其相关不确定性的概率估计。除了其新颖性和理论吸引力外,我们方法的性能与现有近似对数行列式的方法相竞争,同时量化了由于预算受限的证据而导致的不确定性。

英文摘要:

The log-determinant of a kernel matrix appears in a variety of machine learning problems, ranging from determinantal point processes and generalized Markov random fields, through to the training of Gaussian processes. Exact calculation of this term is often intractable when the size of the kernel matrix exceeds a few thousand. In the spirit of probabilistic numerics, we reinterpret the problem of computing the log-determinant as a Bayesian inference problem. In particular, we combine prior knowledge in the form of bounds from matrix theory and evidence derived from stochastic trace estimation to obtain probabilistic estimates for the log-determinant and its associated uncertainty within a given computational budget. Beyond its novelty and theoretic appeal, the performance of our proposal is competitive with state-of-the-art approaches to approximating the log-determinant, while also quantifying the uncertainty due to budget-constrained evidence.

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。

↑