Compressive Spectral Clustering

Authors: Nicolas Tremblay, Gilles Puy, Remi Gribonval, Pierre Vandergheynst

ICML 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We test the performance of our method on artificial and real-world network data.
Researcher Affiliation Collaboration INRIA Rennes Bretagne Atlantique, Campus de Beaulieu, FR-35042 Rennes Cedex, France Institute of Electrical Engineering, Ecole Polytechnique F ed erale de Lausanne (EPFL), CH-1015 Lausanne, Switzerland Technicolor, 975 Avenue des Champs Blancs, 35576 Cesson-S evign e, France
Pseudocode Yes Algorithm 1 Spectral Clustering (Ng et al., 2002)
Open Source Code Yes All our results are reproducible with the CSCbox downloadable at http:// cscbox.gforge.inria.fr/.
Open Datasets Yes We first perform well-controlled experiments on the Stochastic Block Model (SBM)...
Dataset Splits No No explicit details on train/validation/test dataset splits are provided for either the synthetic or real-world datasets.
Hardware Specification Yes Experiments were done on a laptop with a 2.60 GHz Intel i7 dual-core processor running OS Fedora release 22 with 16 GB of RAM.
Software Dependencies Yes Implementation was done in Matlab R2015a... The fast filtering part of CSC uses the gsp cheby op function of the GSP toolbox (Perraudin et al., 2014).
Experiment Setup Yes Input: The Laplacian matrix L, the number of clusters k; and parameters typically set to n = 2k log k, d = 4 log n, p = 50 and γ = 10 3.