Momentum SVGD-EM for Accelerated Maximum Marginal Likelihood Estimation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Rozzio, Adam, Athanasiades, Rafael, Akyildiz, O. Deniz
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917325937049600
author Rozzio, Adam
Athanasiades, Rafael
Akyildiz, O. Deniz
author_facet Rozzio, Adam
Athanasiades, Rafael
Akyildiz, O. Deniz
contents Maximum marginal likelihood estimation (MMLE) can be formulated as the optimization of a free energy functional. From this viewpoint, the Expectation-Maximisation (EM) algorithm admits a natural interpretation as a coordinate descent method over the joint space of model parameters and probability measures. Recently, a significant body of work has adopted this perspective, leading to interacting particle algorithms for MMLE. In this paper, we propose an accelerated version of one such procedure, based on Stein variational gradient descent (SVGD), by introducing Nesterov acceleration in both the parameter updates and in the space of probability measures. The resulting method, termed Momentum SVGD-EM, consistently accelerates convergence in terms of required iterations across various tasks of increasing difficulty, demonstrating effectiveness in both low- and high-dimensional settings.
format Preprint
id arxiv_https___arxiv_org_abs_2603_08676
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Momentum SVGD-EM for Accelerated Maximum Marginal Likelihood Estimation
Rozzio, Adam
Athanasiades, Rafael
Akyildiz, O. Deniz
Machine Learning
Computation
Maximum marginal likelihood estimation (MMLE) can be formulated as the optimization of a free energy functional. From this viewpoint, the Expectation-Maximisation (EM) algorithm admits a natural interpretation as a coordinate descent method over the joint space of model parameters and probability measures. Recently, a significant body of work has adopted this perspective, leading to interacting particle algorithms for MMLE. In this paper, we propose an accelerated version of one such procedure, based on Stein variational gradient descent (SVGD), by introducing Nesterov acceleration in both the parameter updates and in the space of probability measures. The resulting method, termed Momentum SVGD-EM, consistently accelerates convergence in terms of required iterations across various tasks of increasing difficulty, demonstrating effectiveness in both low- and high-dimensional settings.
title Momentum SVGD-EM for Accelerated Maximum Marginal Likelihood Estimation
topic Machine Learning
Computation
url https://arxiv.org/abs/2603.08676