Screening Data Points in Empirical Risk Minimization via Ellipsoidal Regions and Safe Loss Functions - Archive ouverte HAL Access content directly
Conference Papers Year : 2020

Screening Data Points in Empirical Risk Minimization via Ellipsoidal Regions and Safe Loss Functions

Abstract

We design simple screening tests to automatically discard data samples in empirical risk minimization without losing optimization guarantees. We derive loss functions that produce dual objectives with a sparse solution. We also show how to regularize convex losses to ensure such a dual sparsity-inducing property, and propose a general method to design screening tests for classification or regression based on ellipsoidal approximations of the optimal set. In addition to producing computational gains, our approach also allows us to compress a dataset into a subset of representative points.
Fichier principal
Vignette du fichier
main_aistats.pdf (869.18 Ko) Télécharger le fichier
Origin Files produced by the author(s)
Loading...

Dates and versions

hal-02395624 , version 1 (05-12-2019)
hal-02395624 , version 2 (12-06-2020)

Identifiers

  • HAL Id : hal-02395624 , version 2

Cite

Grégoire Mialon, Alexandre d'Aspremont, Julien Mairal. Screening Data Points in Empirical Risk Minimization via Ellipsoidal Regions and Safe Loss Functions. AISTATS 2020 - 23rd International Conference on Artificial Intelligence and Statistics, Aug 2020, Palermo / Virtual, Italy. pp.3610-3620. ⟨hal-02395624v2⟩
2655 View
89 Download

Share

Gmail Mastodon Facebook X LinkedIn More