Stochastic Variance Reduced Optimization for Nonconvex Sparse Learning

Xingguo Li, Tuo Zhao, Raman Arora, Han Liu, Jarvis Haupt
Proceedings of The 33rd International Conference on Machine Learning, PMLR 48:917-925, 2016.

Abstract

We propose a stochastic variance reduced optimization algorithm for solving a class of large-scale nonconvex optimization problems with cardinality constraints, and provide sufficient conditions under which the proposed algorithm enjoys strong linear convergence guarantees and optimal estimation accuracy in high dimensions. Numerical experiments demonstrate the efficiency of our method in terms of both parameter estimation and computational performance.

Cite this Paper


BibTeX
@InProceedings{pmlr-v48-lid16, title = {Stochastic Variance Reduced Optimization for Nonconvex Sparse Learning}, author = {Li, Xingguo and Zhao, Tuo and Arora, Raman and Liu, Han and Haupt, Jarvis}, booktitle = {Proceedings of The 33rd International Conference on Machine Learning}, pages = {917--925}, 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/lid16.pdf}, url = {https://proceedings.mlr.press/v48/lid16.html}, abstract = {We propose a stochastic variance reduced optimization algorithm for solving a class of large-scale nonconvex optimization problems with cardinality constraints, and provide sufficient conditions under which the proposed algorithm enjoys strong linear convergence guarantees and optimal estimation accuracy in high dimensions. Numerical experiments demonstrate the efficiency of our method in terms of both parameter estimation and computational performance.} }
Endnote
%0 Conference Paper %T Stochastic Variance Reduced Optimization for Nonconvex Sparse Learning %A Xingguo Li %A Tuo Zhao %A Raman Arora %A Han Liu %A Jarvis Haupt %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-lid16 %I PMLR %P 917--925 %U https://proceedings.mlr.press/v48/lid16.html %V 48 %X We propose a stochastic variance reduced optimization algorithm for solving a class of large-scale nonconvex optimization problems with cardinality constraints, and provide sufficient conditions under which the proposed algorithm enjoys strong linear convergence guarantees and optimal estimation accuracy in high dimensions. Numerical experiments demonstrate the efficiency of our method in terms of both parameter estimation and computational performance.
RIS
TY - CPAPER TI - Stochastic Variance Reduced Optimization for Nonconvex Sparse Learning AU - Xingguo Li AU - Tuo Zhao AU - Raman Arora AU - Han Liu AU - Jarvis Haupt 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-lid16 PB - PMLR DP - Proceedings of Machine Learning Research VL - 48 SP - 917 EP - 925 L1 - http://proceedings.mlr.press/v48/lid16.pdf UR - https://proceedings.mlr.press/v48/lid16.html AB - We propose a stochastic variance reduced optimization algorithm for solving a class of large-scale nonconvex optimization problems with cardinality constraints, and provide sufficient conditions under which the proposed algorithm enjoys strong linear convergence guarantees and optimal estimation accuracy in high dimensions. Numerical experiments demonstrate the efficiency of our method in terms of both parameter estimation and computational performance. ER -
APA
Li, X., Zhao, T., Arora, R., Liu, H. & Haupt, J.. (2016). Stochastic Variance Reduced Optimization for Nonconvex Sparse Learning. Proceedings of The 33rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 48:917-925 Available from https://proceedings.mlr.press/v48/lid16.html.

Related Material