Quantifying uncertainty in the numerical integration of evolution equations based on Bayesian isotonic regression
Numerical Analysis
2024-11-14 v1 Numerical Analysis
Data Analysis, Statistics and Probability
Methodology
Abstract
This paper presents a new Bayesian framework for quantifying discretization errors in numerical solutions of ordinary differential equations. By modelling the errors as random variables, we impose a monotonicity constraint on the variances, referred to as discretization error variances. The key to our approach is the use of a shrinkage prior for the variances coupled with variable transformations. This methodology extends existing Bayesian isotonic regression techniques to tackle the challenge of estimating the variances of a normal distribution. An additional key feature is the use of a Gaussian mixture model for the - distribution, enabling the development of an efficient Gibbs sampling algorithm for the corresponding posterior.
Cite
@article{arxiv.2411.08338,
title = {Quantifying uncertainty in the numerical integration of evolution equations based on Bayesian isotonic regression},
author = {Yuto Miyatake and Kaoru Irie and Takeru Matsuda},
journal= {arXiv preprint arXiv:2411.08338},
year = {2024}
}