Guardado en:
Detalles Bibliográficos
Autores principales: Vo, Vy, Le, Trung, Vuong, Tung-Long, Zhao, He, Bonilla, Edwin, Phung, Dinh
Formato: Preprint
Publicado: 2023
Materias:
Acceso en línea:https://arxiv.org/abs/2305.15927
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866929368373133312
author Vo, Vy
Le, Trung
Vuong, Tung-Long
Zhao, He
Bonilla, Edwin
Phung, Dinh
author_facet Vo, Vy
Le, Trung
Vuong, Tung-Long
Zhao, He
Bonilla, Edwin
Phung, Dinh
contents Estimating the parameters of a probabilistic directed graphical model from incomplete data is a long-standing challenge. This is because, in the presence of latent variables, both the likelihood function and posterior distribution are intractable without assumptions about structural dependencies or model classes. While existing learning methods are fundamentally based on likelihood maximization, here we offer a new view of the parameter learning problem through the lens of optimal transport. This perspective licenses a general framework that operates on any directed graphs without making unrealistic assumptions on the posterior over the latent variables or resorting to variational approximations. We develop a theoretical framework and support it with extensive empirical evidence demonstrating the versatility and robustness of our approach. Across experiments, we show that not only can our method effectively recover the ground-truth parameters but it also performs comparably or better than competing baselines on downstream applications.
format Preprint
id arxiv_https___arxiv_org_abs_2305_15927
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Parameter Estimation in DAGs from Incomplete Data via Optimal Transport
Vo, Vy
Le, Trung
Vuong, Tung-Long
Zhao, He
Bonilla, Edwin
Phung, Dinh
Machine Learning
Social and Information Networks
Estimating the parameters of a probabilistic directed graphical model from incomplete data is a long-standing challenge. This is because, in the presence of latent variables, both the likelihood function and posterior distribution are intractable without assumptions about structural dependencies or model classes. While existing learning methods are fundamentally based on likelihood maximization, here we offer a new view of the parameter learning problem through the lens of optimal transport. This perspective licenses a general framework that operates on any directed graphs without making unrealistic assumptions on the posterior over the latent variables or resorting to variational approximations. We develop a theoretical framework and support it with extensive empirical evidence demonstrating the versatility and robustness of our approach. Across experiments, we show that not only can our method effectively recover the ground-truth parameters but it also performs comparably or better than competing baselines on downstream applications.
title Parameter Estimation in DAGs from Incomplete Data via Optimal Transport
topic Machine Learning
Social and Information Networks
url https://arxiv.org/abs/2305.15927