Bayesian sequential parameter estimation with a Laplace type approximation

TL;DR

This paper introduces a sequential inference method for dynamic latent Gaussian models using iterated Laplace approximation, balancing computational efficiency and accuracy, with improvements from correction techniques and a population-based approach.

Contribution

It presents a novel sequential inference technique for latent Gaussian models leveraging iterated Laplace approximation, with enhancements for robustness and accuracy.

Findings

Approximation corrections improve posterior accuracy.

Population-based approach enhances robustness.

Method balances computational efficiency and inference precision.

Abstract

A method for sequential inference of the fixed parameters of a dynamic latent Gaussian models is proposed and evaluated that is based on the iterated Laplace approximation. The method provides a useful trade-off between computational performance and the accuracy of the approximation to the true posterior distribution. Approximation corrections are shown to improve the accuracy of the approximation in simulation studies. A population-based approach is also shown to provide a more robust inference method.