Sketching Datasets for Large-Scale Learning (long version)

Rémi Gribonval (1, 2) , Antoine Chatalic (1) , Nicolas Keriven (3) , Vincent Schellekens (4) , Laurent Jacques (4) , Philip Schniter (5)

1 PANAMA - Parcimonie et Nouveaux Algorithmes pour le Signal et la Modélisation Audio
2 DANTE - Dynamic Networks : Temporal and Structural Capture Approach
3 GIPSA-GAIA - GIPSA Pôle Géométrie, Apprentissage, Information et Algorithmes
4 ICTEAM - Institute of Information and Communication Technologies, Electronics and Applied Mathematics
5 OSU - Ohio State University [Columbus]

This article considers "sketched learning," or "compressive learning," an approach to large-scale machine learning where datasets are massively compressed before learning (e.g., clustering, classification, or regression) is performed. In particular, a "sketch" is first constructed by computing carefully chosen nonlinear random features (e.g., random Fourier features) and averaging them over the whole dataset. Parameters are then learned from the sketch, without access to the original dataset. This article surveys the current state-of-the-art in sketched learning, including the main concepts and algorithms, their connections with established signal-processing methods, existing theoretical guarantees-on both information preservation and privacy preservation, and important open problems.

Domaines: Traitement du signal et de l'image [eess.SP]; Apprentissage [cs.LG]

Rémi Gribonval, Antoine Chatalic, Nicolas Keriven, Vincent Schellekens, Laurent Jacques, et al.. Sketching Datasets for Large-Scale Learning (long version). 2021. ⟨hal-02909766v2⟩