Header logo is

A probabilistic model for the numerical solution of initial value problems




We study connections between ordinary differential equation (ODE) solvers and probabilistic regression methods in statistics. We provide a new view of probabilistic ODE solvers as active inference agents operating on stochastic differential equation models that estimate the unknown initial value problem (IVP) solution from approximate observations of the solution derivative, as provided by the ODE dynamics. Adding to this picture, we show that several multistep methods of Nordsieck form can be recast as Kalman filtering on q-times integrated Wiener processes. Doing so provides a family of IVP solvers that return a Gaussian posterior measure, rather than a point estimate. We show that some such methods have low computational overhead, nontrivial convergence order, and that the posterior has a calibrated concentration rate. Additionally, we suggest a step size adaptation algorithm which completes the proposed method to a practically useful implementation, which we experimentally evaluate using a representative set of standard codes in the DETEST benchmark set.

Author(s): Michael Schober and Simon Särkkä and Philipp Hennig,
Journal: Statistics and Computing
Year: 2018
Publisher: Springer US

Department(s): Probabilistic Numerics
Research Project(s): Probabilistic Solvers for Ordinary Differential Equations
Bibtex Type: Article (article)
Paper Type: Journal

DOI: 10.1007/s11222-017-9798-7

Links: PDF


  title = {A probabilistic model for the numerical solution of initial value problems},
  author = {Schober, Michael and S{\"a}rkk{\"a}, Simon and Philipp Hennig},
  journal = {Statistics and Computing},
  publisher = {Springer US},
  year = {2018}