Discrete gradients for computational Bayesian inference

03/01/2019
by   Sahani Pathiraja, et al.
0

In this paper, we exploit the gradient flow structure of continuous-time formulations of Bayesian inference in terms of their numerical time-stepping. We focus on two particular examples, namely, the continuous-time ensemble Kalman-Bucy filter and a particle discretisation of the Fokker-Planck equation associated to Brownian dynamics. Both formulations can lead to stiff differential equations which require special numerical methods for their efficient numerical implementation. We compare discrete gradient methods to alternative semi-implicit and other iterative implementations of the underlying Bayesian inference problems.

READ FULL TEXT
research
11/25/2019

Fokker-Planck particle systems for Bayesian inference: Computational approaches

Bayesian inference can be embedded into an appropriately defined dynamic...
research
09/23/2022

Ensemble-based gradient inference for particle methods in optimization and sampling

We propose an approach based on function evaluations and Bayesian infere...
research
03/21/2019

The Hitchhikers Guide to Nonlinear Filtering

Nonlinear filtering is the problem of online estimation of a dynamic hid...
research
11/20/2015

Bayesian inference via rejection filtering

We provide a method for approximating Bayesian inference using rejection...
research
06/22/2021

Distributional Gradient Matching for Learning Uncertain Neural Dynamics Models

Differential equations in general and neural ODEs in particular are an e...
research
12/15/2022

Time-limited Balanced Truncation for Data Assimilation Problems

Balanced truncation is a well-established model order reduction method i...
research
04/25/2023

On forward-backward SDE approaches to continuous-time minimum variance estimation

The work of Kalman and Bucy has established a duality between filtering ...

Please sign up or login with your details

Forgot password? Click here to reset