Proximal interacting particle langevin algorithms
File(s)
Accepted version
Author(s)
Cordero Encinar, paula
Crucinio, Francesca
Akyildiz, Omer Deniz
Type
Conference Paper
Abstract
We introduce a class of algorithms, termed proximal interacting particle Langevin algorithms (PIPLA), for inference and learning in latent variable models whose joint probability density is non-differentiable. Leveraging proximal Markov chain Monte Carlo techniques and interacting particle Langevin algorithms, we propose three algorithms tailored to the problem of estimating parameters in a non-differentiable statistical model. We prove non-asymptotic bounds for the parameter estimates produced by the different algorithms in the strongly log-concave setting and provide comprehensive numerical experiments on various models to demonstrate the effectiveness of the proposed methods. In particular, we demonstrate the utility of our family of algorithms for sparse Bayesian logistic regression, training of sparse Bayesian neural networks or neural networks with non-differentiable activation functions, image deblurring, and sparse matrix completion. Our theory and experiments together show that PIPLA family can be the de facto choice for parameter estimation problems in non-differentiable latent variable models.
Date Issued
2025-07-22
Date Acceptance
2025-05-07
Citation
Proceedings of Machine Learning Research, 2025, 286, pp.1220-1265
ISSN
2640-3498
Publisher
MLResearchPress
Start Page
1220
End Page
1265
Journal / Book Title
Proceedings of Machine Learning Research
Volume
286
Copyright Statement
This paper is embargoed until publication.
Source
Uncertainty in Artificial Intelligence (UAI) 2025
Publication Status
Published
Start Date
2025-07-22
Finish Date
2025-07-24
Coverage Spatial
Rio de Janeiro, Brazil