Solving Ridge Regression using Sketched Preconditioned SVRG

Alon Gonen, Francesco Orabona, Shai Shalev-Shwartz
Proceedings of The 33rd International Conference on Machine Learning, PMLR 48:1397-1405, 2016.

Abstract

We develop a novel preconditioning method for ridge regression, based on recent linear sketching methods. By equipping Stochastic Variance Reduced Gradient (SVRG) with this preconditioning process, we obtain a significant speed-up relative to fast stochastic methods such as SVRG, SDCA and SAG.

Cite this Paper

BibTeX
@InProceedings{pmlr-v48-gonen16, title = {Solving Ridge Regression using Sketched Preconditioned SVRG}, author = {Gonen, Alon and Orabona, Francesco and Shalev-Shwartz, Shai}, booktitle = {Proceedings of The 33rd International Conference on Machine Learning}, pages = {1397--1405}, year = {2016}, editor = {Balcan, Maria Florina and Weinberger, Kilian Q.}, volume = {48}, series = {Proceedings of Machine Learning Research}, address = {New York, New York, USA}, month = {20--22 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v48/gonen16.pdf}, url = {https://proceedings.mlr.press/v48/gonen16.html}, abstract = {We develop a novel preconditioning method for ridge regression, based on recent linear sketching methods. By equipping Stochastic Variance Reduced Gradient (SVRG) with this preconditioning process, we obtain a significant speed-up relative to fast stochastic methods such as SVRG, SDCA and SAG.} }
Endnote
%0 Conference Paper %T Solving Ridge Regression using Sketched Preconditioned SVRG %A Alon Gonen %A Francesco Orabona %A Shai Shalev-Shwartz %B Proceedings of The 33rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2016 %E Maria Florina Balcan %E Kilian Q. Weinberger %F pmlr-v48-gonen16 %I PMLR %P 1397--1405 %U https://proceedings.mlr.press/v48/gonen16.html %V 48 %X We develop a novel preconditioning method for ridge regression, based on recent linear sketching methods. By equipping Stochastic Variance Reduced Gradient (SVRG) with this preconditioning process, we obtain a significant speed-up relative to fast stochastic methods such as SVRG, SDCA and SAG.
RIS
TY - CPAPER TI - Solving Ridge Regression using Sketched Preconditioned SVRG AU - Alon Gonen AU - Francesco Orabona AU - Shai Shalev-Shwartz BT - Proceedings of The 33rd International Conference on Machine Learning DA - 2016/06/11 ED - Maria Florina Balcan ED - Kilian Q. Weinberger ID - pmlr-v48-gonen16 PB - PMLR DP - Proceedings of Machine Learning Research VL - 48 SP - 1397 EP - 1405 L1 - http://proceedings.mlr.press/v48/gonen16.pdf UR - https://proceedings.mlr.press/v48/gonen16.html AB - We develop a novel preconditioning method for ridge regression, based on recent linear sketching methods. By equipping Stochastic Variance Reduced Gradient (SVRG) with this preconditioning process, we obtain a significant speed-up relative to fast stochastic methods such as SVRG, SDCA and SAG. ER -
APA
Gonen, A., Orabona, F. & Shalev-Shwartz, S.. (2016). Solving Ridge Regression using Sketched Preconditioned SVRG. Proceedings of The 33rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 48:1397-1405 Available from https://proceedings.mlr.press/v48/gonen16.html.

Related Material