Saved in:
Bibliographic Details
Main Authors: Olmin, Amanda, Lindqvist, Jakob, Svensson, Lennart, Lindsten, Fredrik
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2402.16688
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929256810938368
author Olmin, Amanda
Lindqvist, Jakob
Svensson, Lennart
Lindsten, Fredrik
author_facet Olmin, Amanda
Lindqvist, Jakob
Svensson, Lennart
Lindsten, Fredrik
contents Noise-contrastive estimation (NCE) is a popular method for estimating unnormalised probabilistic models, such as energy-based models, which are effective for modelling complex data distributions. Unlike classical maximum likelihood (ML) estimation that relies on importance sampling (resulting in ML-IS) or MCMC (resulting in contrastive divergence, CD), NCE uses a proxy criterion to avoid the need for evaluating an often intractable normalisation constant. Despite apparent conceptual differences, we show that two NCE criteria, ranking NCE (RNCE) and conditional NCE (CNCE), can be viewed as ML estimation methods. Specifically, RNCE is equivalent to ML estimation combined with conditional importance sampling, and both RNCE and CNCE are special cases of CD. These findings bridge the gap between the two method classes and allow us to apply techniques from the ML-IS and CD literature to NCE, offering several advantageous extensions.
format Preprint
id arxiv_https___arxiv_org_abs_2402_16688
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle On the connection between Noise-Contrastive Estimation and Contrastive Divergence
Olmin, Amanda
Lindqvist, Jakob
Svensson, Lennart
Lindsten, Fredrik
Machine Learning
Noise-contrastive estimation (NCE) is a popular method for estimating unnormalised probabilistic models, such as energy-based models, which are effective for modelling complex data distributions. Unlike classical maximum likelihood (ML) estimation that relies on importance sampling (resulting in ML-IS) or MCMC (resulting in contrastive divergence, CD), NCE uses a proxy criterion to avoid the need for evaluating an often intractable normalisation constant. Despite apparent conceptual differences, we show that two NCE criteria, ranking NCE (RNCE) and conditional NCE (CNCE), can be viewed as ML estimation methods. Specifically, RNCE is equivalent to ML estimation combined with conditional importance sampling, and both RNCE and CNCE are special cases of CD. These findings bridge the gap between the two method classes and allow us to apply techniques from the ML-IS and CD literature to NCE, offering several advantageous extensions.
title On the connection between Noise-Contrastive Estimation and Contrastive Divergence
topic Machine Learning
url https://arxiv.org/abs/2402.16688