PENEX: AdaBoost-Inspired Neural Network Regularization

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kladny, Klaus-Rudolf, Schölkopf, Bernhard, Muehlebach, Michael
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913114698547200
author Kladny, Klaus-Rudolf
Schölkopf, Bernhard
Muehlebach, Michael
author_facet Kladny, Klaus-Rudolf
Schölkopf, Bernhard
Muehlebach, Michael
contents AdaBoost sequentially fits so-called weak learners to minimize an exponential loss, which penalizes misclassified data points more severely than other loss functions like cross-entropy. Paradoxically, AdaBoost generalizes well in practice as the number of weak learners grows. In the present work, we introduce Penalized Exponential Loss (PENEX), a new formulation of the multi-class exponential loss that is theoretically grounded and, in contrast to the existing formulation, amenable to optimization via first-order methods, making it a practical objective for training neural networks. We demonstrate that PENEX effectively increases margins of data points, which can be translated into a generalization bound. Empirically, across computer vision and language tasks, PENEX improves neural network generalization in low-data regimes, matching and in some settings outperforming established regularizers at comparable computational cost. Our results highlight the potential of the exponential loss beyond its application in AdaBoost.
format Preprint
id arxiv_https___arxiv_org_abs_2510_02107
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle PENEX: AdaBoost-Inspired Neural Network Regularization
Kladny, Klaus-Rudolf
Schölkopf, Bernhard
Muehlebach, Michael
Machine Learning
AdaBoost sequentially fits so-called weak learners to minimize an exponential loss, which penalizes misclassified data points more severely than other loss functions like cross-entropy. Paradoxically, AdaBoost generalizes well in practice as the number of weak learners grows. In the present work, we introduce Penalized Exponential Loss (PENEX), a new formulation of the multi-class exponential loss that is theoretically grounded and, in contrast to the existing formulation, amenable to optimization via first-order methods, making it a practical objective for training neural networks. We demonstrate that PENEX effectively increases margins of data points, which can be translated into a generalization bound. Empirically, across computer vision and language tasks, PENEX improves neural network generalization in low-data regimes, matching and in some settings outperforming established regularizers at comparable computational cost. Our results highlight the potential of the exponential loss beyond its application in AdaBoost.
title PENEX: AdaBoost-Inspired Neural Network Regularization
topic Machine Learning
url https://arxiv.org/abs/2510.02107