Contrastive Learning with Nasty Noise

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autor principal: Zhao, Ziruo
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866912244438138880
author Zhao, Ziruo
author_facet Zhao, Ziruo
contents Contrastive learning has emerged as a powerful paradigm for self-supervised representation learning. This work analyzes the theoretical limits of contrastive learning under nasty noise, where an adversary modifies or replaces training samples. Using PAC learning and VC-dimension analysis, lower and upper bounds on sample complexity in adversarial settings are established. Additionally, data-dependent sample complexity bounds based on the l2-distance function are derived.
format Preprint
id arxiv_https___arxiv_org_abs_2502_17872
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Contrastive Learning with Nasty Noise
Zhao, Ziruo
Machine Learning
Artificial Intelligence
Contrastive learning has emerged as a powerful paradigm for self-supervised representation learning. This work analyzes the theoretical limits of contrastive learning under nasty noise, where an adversary modifies or replaces training samples. Using PAC learning and VC-dimension analysis, lower and upper bounds on sample complexity in adversarial settings are established. Additionally, data-dependent sample complexity bounds based on the l2-distance function are derived.
title Contrastive Learning with Nasty Noise
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2502.17872