Spectral Properties of Radial Kernels and Clustering in High Dimensions - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Preprints, Working Papers, ... Year : 2020

Spectral Properties of Radial Kernels and Clustering in High Dimensions

Abstract

In this paper, we study the spectrum and the eigenvectors of radial kernels for mixtures of distributions in R n. Our approach focuses on high dimensions and relies solely on the concentration properties of the components in the mixture. We give several results describing of the structure of kernel matrices for a sample drawn from such a mixture. Based on these results, we analyze the ability of kernel PCA to cluster high dimensional mixtures. In particular, we exhibit a specific kernel leading to a simple spectral algorithm for clustering mixtures with possibly common means but different covariance matrices. We show that the minimum angular separation between the covariance matrices that is required for the algorithm to succeed tends to 0 as n goes to infinity.
Fichier principal
Vignette du fichier
hdkernel.pdf (395.93 Ko) Télécharger le fichier
Origin Files produced by the author(s)
Loading...

Dates and versions

hal-01969956 , version 1 (04-01-2019)
hal-01969956 , version 2 (23-01-2019)
hal-01969956 , version 3 (15-03-2019)
hal-01969956 , version 4 (25-06-2019)
hal-01969956 , version 5 (27-06-2019)
hal-01969956 , version 6 (28-06-2019)
hal-01969956 , version 7 (06-01-2020)

Identifiers

  • HAL Id : hal-01969956 , version 7

Cite

David Cohen-Steiner, Alba Chiara de Vitis. Spectral Properties of Radial Kernels and Clustering in High Dimensions. 2020. ⟨hal-01969956v7⟩
357 View
378 Download

Share

Gmail Mastodon Facebook X LinkedIn More