Nyström method with Kernel K-means++ samples as landmarks

We investigate, theoretically and empirically, the effectiveness of kernel K-means++ samples as landmarks in the Nyström method for low-rank approximation of kernel matrices. Previous empirical studies (Zhang et al., 2008; Kumar et al.,2012) observe that the landmarks obtained using (kernel) K-means...

Full description

Bibliographic Details
Main Authors: Oglic, Dino, Gaertner, Thomas
Format: Conference or Workshop Item
Published: Journal of Machine Learning Research 2017
Online Access:https://eprints.nottingham.ac.uk/43573/