Large Scale Kernel Learning using Block Coordinate Descent
Machine Learning
2016-02-18 v1 Optimization and Control
Machine Learning
Abstract
We demonstrate that distributed block coordinate descent can quickly solve kernel regression and classification problems with millions of data points. Armed with this capability, we conduct a thorough comparison between the full kernel, the Nystr\"om method, and random features on three large classification tasks from various domains. Our results suggest that the Nystr\"om method generally achieves better statistical accuracy than random features, but can require significantly more iterations of optimization. Lastly, we derive new rates for block coordinate descent which support our experimental findings when specialized to kernel methods.
Cite
@article{arxiv.1602.05310,
title = {Large Scale Kernel Learning using Block Coordinate Descent},
author = {Stephen Tu and Rebecca Roelofs and Shivaram Venkataraman and Benjamin Recht},
journal= {arXiv preprint arXiv:1602.05310},
year = {2016}
}