Accueil

Filtering and sampling of graph signals, and its application to clustering

Mercredi 13 juin 10:30-11:30 - Nicolas Tremblay - Laboratoire GIPSA, Grenoble

Filtering and sampling of graph signals, and its application to clustering

Résumé : Spectral clustering has become a popular technique due to its high performance in many contexts. It comprises three main steps : create a similarity graph between N objects to cluster, compute the first k eigenvectors of its Laplacian matrix to define a feature vector for each object, and run k-means on these features to separate objects into k classes. Each of these three steps becomes computationally intensive for large N and/or k. We propose to speed up the last two steps based on recent results in the emerging field of graph signal processing : graph filtering of random signals, and random sampling of bandlimited graph signals. In this presentation, we will take time to go over what filtering and sampling mean for a signal defined on a graph, and explain to what extent they can prove useful for spectral clustering.

Pour en savoir plus sur cet événement, consultez l'article Séminaires SPOC