paper-with-me

Papers

On the connection between Noise-Contrastive Estimation and Contrastive Divergence

2024-02-26 · Amanda Olmin, Jakob Lindqvist, Lennart Svensson, Fredrik Lindsten

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.

📄 PDF Abstract BibTeX arXiv:2402.16688

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Data Augmentation of Contrastive Learning is Estimating Positive-incentive Noise

2024-08-19 · Hongyuan Zhang, Yanchen Xu, Sida Huang, Xuelong Li

Inspired by the idea of Positive-incentive Noise (Pi-Noise or $\pi$-Noise) that aims at learning the reliable noise beneficial to tasks, we scientifically investigate the connection between contrastive learning and $\pi$…

Contrastive LearningData Augmentation

Your contrastive learning problem is secretly a distribution alignment problem

2025-02-27 · Zihao Chen, Chi-Heng Lin, Ran Liu, Jingyun Xiao 외

Despite the success of contrastive learning (CL) in vision and language, its theoretical foundations and mechanisms for building representations remain poorly understood. In this work, we build connections between noise …

Contrastive LearningSelf-Supervised Learning

From $t$-SNE to UMAP with contrastive learning

2022-06-03 · Sebastian Damrich, Jan Niklas Böhm, Fred A. Hamprecht, Dmitry Kobak

Neighbor embedding methods $t$-SNE and UMAP are the de facto standard for visualizing high-dimensional datasets. Motivated from entirely different viewpoints, their loss functions appear to be unrelated. In practice, the…

Contrastive LearningRepresentation Learning

Similarity Contrastive Estimation for Image and Video Soft Contrastive Self-Supervised Learning

2022-12-21 · Julien Denize, Jaonary Rabarisoa, Astrid Orcesi, Romain Hérault

Contrastive representation learning has proven to be an effective self-supervised learning method for images and videos. Most successful approaches are based on Noise Contrastive Estimation (NCE) and use different views …

Contrastive LearningLinear evaluationRepresentation LearningSelf-Supervised Action Recognition+4

Similarity Contrastive Estimation for Self-Supervised Soft Contrastive Learning

2021-11-29 · Julien Denize, Jaonary Rabarisoa, Astrid Orcesi, Romain Hérault 외

Contrastive representation learning has proven to be an effective self-supervised learning method. Most successful approaches are based on Noise Contrastive Estimation (NCE) and use different views of an instance as posi…

Contrastive LearningLinear evaluationRepresentation LearningSelf-Supervised Image Classification+3