arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1306.0308stat.MLcs.LGcs.NAmath.NA
收藏

微分方程的概率解及其在黎曼统计学中的应用

Probabilistic Solutions to Differential Equations and their Application to Riemannian Statistics

Max Planck Institute for Intelligent Systems(马克斯·普朗克智能系统研究所)

Philipp Hennig, Søren Hauberg

AI总结:

本文提出一种概率数值方法,用于求解初值和边界值问题,返回解的高斯过程后验。该方法在黎曼流形统计中具有应用价值,能处理非解析常微分方程,通过不确定性边际化提升统计鲁棒性,提出新的黎曼算法和主地理分析方法。

摘要:

我们研究了一种概率数值方法,用于求解初值和边界值问题,该方法返回解的联合高斯过程后验。此类方法在黎曼流形统计中具有实际价值,因为几乎所有计算都涉及非解析常微分方程。概率公式允许对数值解的不确定性进行边际化,使统计结果对不准确性更不敏感。这导致了新的黎曼算法用于均值计算和主地理分析。边际化也意味着结果可能不如点估计精确,从而在状态-of-the-art方法上实现显著加速。我们的方法是关于更广泛观点的论据,即数值计算引起的不确定性应在机器学习算法的整个管道中进行跟踪。

英文摘要:

We study a probabilistic numerical method for the solution of both boundary and initial value problems that returns a joint Gaussian process posterior over the solution. Such methods have concrete value in the statistics on Riemannian manifolds, where non-analytic ordinary differential equations are involved in virtually all computations. The probabilistic formulation permits marginalising the uncertainty of the numerical solution such that statistics are less sensitive to inaccuracies. This leads to new Riemannian algorithms for mean value computations and principal geodesic analysis. Marginalisation also means results can be less precise than point estimates, enabling a noticeable speed-up over the state of the art. Our approach is an argument for a wider point that uncertainty caused by numerical calculations should be tracked throughout the pipeline of machine learning algorithms.

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。

↑