Black Box Probabilistic Numerics

TL;DR

This paper introduces a black box probabilistic numerics approach that leverages only the final output of traditional methods to solve complex numerical tasks, broadening applicability and improving convergence.

Contribution

It proposes a novel black box framework for probabilistic numerics that uses only endpoint data, enabling higher-order convergence and application to nonlinear differential and eigenvalue problems.

Findings

Expands probabilistic numerics to new problem classes

Achieves higher orders of convergence

Demonstrates effectiveness on differential and eigenvalue problems

Abstract

Probabilistic numerics casts numerical tasks, such the numerical solution of differential equations, as inference problems to be solved. One approach is to model the unknown quantity of interest as a random variable, and to constrain this variable using data generated during the course of a traditional numerical method. However, data may be nonlinearly related to the quantity of interest, rendering the proper conditioning of random variables difficult and limiting the range of numerical tasks that can be addressed. Instead, this paper proposes to construct probabilistic numerical methods based only on the final output from a traditional method. A convergent sequence of approximations to the quantity of interest constitute a dataset, from which the limiting quantity of interest can be extrapolated, in a probabilistic analogue of Richardson's deferred approach to the limit. This black box approach (1) massively expands the range of tasks to which probabilistic numerics can be applied, (2) inherits the features and performance of state-of-the-art numerical methods, and (3) enables provably higher orders of convergence to be achieved. Applications are presented for nonlinear ordinary and partial differential equations, as well as for eigenvalue problems-a setting for which no probabilistic numerical methods have yet been developed.