Particle algorithms for maximum likelihood training of latent variable models

Author name not available (Why is that?)

Publication date: 27 April 2022

Abstract: (Neal and Hinton, 1998) recast maximum likelihood estimation of any given latent variable model as the minimization of a free energy functional

F

, and the EM algorithm as coordinate descent applied to

F

. Here, we explore alternative ways to optimize the functional. In particular, we identify various gradient flows associated with

F

and show that their limits coincide with

F

's stationary points. By discretizing the flows, we obtain practical particle-based algorithms for maximum likelihood estimation in broad classes of latent variable models. The novel algorithms scale to high-dimensional settings and perform well in numerical experiments.

Has companion code repository: https://github.com/juankuntz/parem

This page was built for publication: Particle algorithms for maximum likelihood training of latent variable models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6397617)