Saved in:
Bibliographic Details
Main Authors: Millard, Andrew, Murphy, Joshua, Green, Peter, Maskell, Simon
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2601.21983
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918313642164224
author Millard, Andrew
Murphy, Joshua
Green, Peter
Maskell, Simon
author_facet Millard, Andrew
Murphy, Joshua
Green, Peter
Maskell, Simon
contents Bayesian inference allows us to define a posterior distribution over the weights of a generic neural network (NN). Exact posteriors are usually intractable, in which case approximations can be employed. One such approximation - variational inference - is computationally efficient when using mini-batch stochastic gradient descent as subsets of the data are used for likelihood and gradient evaluations, though the approach relies on the selection of a variational distribution which sufficiently matches the form of the posterior. Particle-based methods such as Markov chain Monte Carlo and Sequential Monte Carlo (SMC) do not assume a parametric family for the posterior by typically require higher computational cost. These sampling methods typically use the full-batch of data for likelihood and gradient evaluations, which contributes to this computational expense. We explore several methods of gradually introducing more mini-batches of data (data annealing) into likelihood and gradient evaluations of an SMC sampler. We find that we can achieve up to $6\times$ faster training with minimal loss in accuracy on benchmark image classification problems using NNs.
format Preprint
id arxiv_https___arxiv_org_abs_2601_21983
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Investigating Batch Inference in a Sequential Monte Carlo Framework for Neural Networks
Millard, Andrew
Murphy, Joshua
Green, Peter
Maskell, Simon
Machine Learning
Bayesian inference allows us to define a posterior distribution over the weights of a generic neural network (NN). Exact posteriors are usually intractable, in which case approximations can be employed. One such approximation - variational inference - is computationally efficient when using mini-batch stochastic gradient descent as subsets of the data are used for likelihood and gradient evaluations, though the approach relies on the selection of a variational distribution which sufficiently matches the form of the posterior. Particle-based methods such as Markov chain Monte Carlo and Sequential Monte Carlo (SMC) do not assume a parametric family for the posterior by typically require higher computational cost. These sampling methods typically use the full-batch of data for likelihood and gradient evaluations, which contributes to this computational expense. We explore several methods of gradually introducing more mini-batches of data (data annealing) into likelihood and gradient evaluations of an SMC sampler. We find that we can achieve up to $6\times$ faster training with minimal loss in accuracy on benchmark image classification problems using NNs.
title Investigating Batch Inference in a Sequential Monte Carlo Framework for Neural Networks
topic Machine Learning
url https://arxiv.org/abs/2601.21983