Learning Noisy Halfspaces with a Margin: Massart is No Harder than Random

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Chandrasekaran, Gautam, Kontonis, Vasilis, Stavropoulos, Konstantinos, Tian, Kevin
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866912191919161344
author Chandrasekaran, Gautam
Kontonis, Vasilis
Stavropoulos, Konstantinos
Tian, Kevin
author_facet Chandrasekaran, Gautam
Kontonis, Vasilis
Stavropoulos, Konstantinos
Tian, Kevin
contents We study the problem of PAC learning $γ$-margin halfspaces with Massart noise. We propose a simple proper learning algorithm, the Perspectron, that has sample complexity $\widetilde{O}((εγ)^{-2})$ and achieves classification error at most $η+ε$ where $η$ is the Massart noise rate. Prior works [DGT19,CKMY20] came with worse sample complexity guarantees (in both $ε$ and $γ$) or could only handle random classification noise [DDK+23,KIT+23] -- a much milder noise assumption. We also show that our results extend to the more challenging setting of learning generalized linear models with a known link function under Massart noise, achieving a similar sample complexity to the halfspace case. This significantly improves upon the prior state-of-the-art in this setting due to [CKMY20], who introduced this model.
format Preprint
id arxiv_https___arxiv_org_abs_2501_09851
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Learning Noisy Halfspaces with a Margin: Massart is No Harder than Random
Chandrasekaran, Gautam
Kontonis, Vasilis
Stavropoulos, Konstantinos
Tian, Kevin
Machine Learning
Data Structures and Algorithms
We study the problem of PAC learning $γ$-margin halfspaces with Massart noise. We propose a simple proper learning algorithm, the Perspectron, that has sample complexity $\widetilde{O}((εγ)^{-2})$ and achieves classification error at most $η+ε$ where $η$ is the Massart noise rate. Prior works [DGT19,CKMY20] came with worse sample complexity guarantees (in both $ε$ and $γ$) or could only handle random classification noise [DDK+23,KIT+23] -- a much milder noise assumption. We also show that our results extend to the more challenging setting of learning generalized linear models with a known link function under Massart noise, achieving a similar sample complexity to the halfspace case. This significantly improves upon the prior state-of-the-art in this setting due to [CKMY20], who introduced this model.
title Learning Noisy Halfspaces with a Margin: Massart is No Harder than Random
topic Machine Learning
Data Structures and Algorithms
url https://arxiv.org/abs/2501.09851