SketchyCoreSVD: SketchySVD from Random Subsampling of the Data Matrix

Abstract

We present a method called SketchyCoreSVD to compute the near-optimal rank r SVD of a data matrix by building random sketches only from its subsampled columns and rows. We provide theoretical guarantees under incoherence assumptions, and validate the performance of our SketchyCoreSVD method on various large static and time-varying datasets.

References in corpus (2)

• The spectral norm error of the naive Nystrom extension
• Matrix Coherence and the Nystrom Method