Skip to Main content Skip to Navigation
Conference papers

Accelerated spectral clustering using graph filtering of random signals

Abstract : We build upon recent advances in graph signal processing to propose a faster spectral clustering algorithm. Indeed, classical spectral clustering is based on the computation of the first k eigenvectors of the similarity matrix' Laplacian, whose computation cost, even for sparse matrices, becomes prohibitive for large datasets. We show that we can estimate the spectral clustering distance matrix without computing these eigenvectors: by graph filtering random signals. Also, we take advantage of the stochasticity of these random vectors to estimate the number of clusters k. We compare our method to classical spectral clustering on synthetic data, and show that it reaches equal performance while being faster by a factor at least two for large datasets.
Document type :
Conference papers
Complete list of metadatas

Cited literature [23 references]  Display  Hide  Download
Contributor : Nicolas Tremblay <>
Submitted on : Tuesday, January 12, 2016 - 3:40:56 PM
Last modification on : Tuesday, December 8, 2020 - 9:50:00 AM
Long-term archiving on: : Friday, November 11, 2016 - 1:50:12 AM


ICASSP_2016_lap (1).pdf
Files produced by the author(s)



Nicolas Tremblay, Gilles Puy, Pierre Borgnat, Rémi Gribonval, Pierre Vandergheynst. Accelerated spectral clustering using graph filtering of random signals. 41st IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2016), Mar 2016, Shanghai, China. ⟨hal-01243682v2⟩



Record views


Files downloads