On the Spectrum of Random Features Maps of High Dimensional Data

Zhenyu Liao 1 Romain Couillet 1, 2
2 GIPSA-CICS - CICS
GIPSA-DIS - Département Images et Signal
Abstract : Random feature maps are ubiquitous in modern statistical machine learning, where they generalize random projections by means of powerful, yet often difficult to analyze nonlinear operators. In this paper, we leverage the "concentration" phenomenon induced by random matrix theory to perform a spectral analysis on the Gram matrix of these random feature maps, here for Gaussian mixture models of simultaneously large dimension and size. Our results are instrumental to a deeper understanding on the interplay of the nonlinearity and the statistics of the data, thereby allowing for a better tuning of random feature-based techniques.
Type de document :
Communication dans un congrès
International Conference on Machine Learning (ICML 2018), Jul 2018, Stockholm, Sweden
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-01954933
Contributeur : Zhenyu Liao <>
Soumis le : vendredi 14 décembre 2018 - 09:24:52
Dernière modification le : mercredi 13 mars 2019 - 13:53:09

Lien texte intégral

Identifiants

  • HAL Id : hal-01954933, version 1
  • ARXIV : 1805.11916

Citation

Zhenyu Liao, Romain Couillet. On the Spectrum of Random Features Maps of High Dimensional Data. International Conference on Machine Learning (ICML 2018), Jul 2018, Stockholm, Sweden. 〈hal-01954933〉

Partager

Métriques

Consultations de la notice

83