Robustly Learning Monotone Generalized Linear Models via Data Augmentation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zarifis, Nikos, Wang, Puqian, Diakonikolas, Ilias, Diakonikolas, Jelena
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915424543703040
author Zarifis, Nikos
Wang, Puqian
Diakonikolas, Ilias
Diakonikolas, Jelena
author_facet Zarifis, Nikos
Wang, Puqian
Diakonikolas, Ilias
Diakonikolas, Jelena
contents We study the task of learning Generalized Linear models (GLMs) in the agnostic model under the Gaussian distribution. We give the first polynomial-time algorithm that achieves a constant-factor approximation for \textit{any} monotone Lipschitz activation. Prior constant-factor GLM learners succeed for a substantially smaller class of activations. Our work resolves a well-known open problem, by developing a robust counterpart to the classical GLMtron algorithm (Kakade et al., 2011). Our robust learner applies more generally, encompassing all monotone activations with bounded $(2+ζ)$-moments, for any fixed $ζ>0$ -- a condition that is essentially necessary. To obtain our results, we leverage a novel data augmentation technique with decreasing Gaussian noise injection and prove a number of structural results that may be useful in other settings.
format Preprint
id arxiv_https___arxiv_org_abs_2502_08611
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Robustly Learning Monotone Generalized Linear Models via Data Augmentation
Zarifis, Nikos
Wang, Puqian
Diakonikolas, Ilias
Diakonikolas, Jelena
Machine Learning
Optimization and Control
Statistics Theory
We study the task of learning Generalized Linear models (GLMs) in the agnostic model under the Gaussian distribution. We give the first polynomial-time algorithm that achieves a constant-factor approximation for \textit{any} monotone Lipschitz activation. Prior constant-factor GLM learners succeed for a substantially smaller class of activations. Our work resolves a well-known open problem, by developing a robust counterpart to the classical GLMtron algorithm (Kakade et al., 2011). Our robust learner applies more generally, encompassing all monotone activations with bounded $(2+ζ)$-moments, for any fixed $ζ>0$ -- a condition that is essentially necessary. To obtain our results, we leverage a novel data augmentation technique with decreasing Gaussian noise injection and prove a number of structural results that may be useful in other settings.
title Robustly Learning Monotone Generalized Linear Models via Data Augmentation
topic Machine Learning
Optimization and Control
Statistics Theory
url https://arxiv.org/abs/2502.08611