Safe Subspace Screening for Nuclear Norm Regularized Least Squares Problems

Qiang Zhou, Qi Zhao
Proceedings of the 32nd International Conference on Machine Learning, PMLR 37:1103-1112, 2015.

Abstract

Nuclear norm regularization has been shown very promising for pursing a low rank matrix solution in various machine learning problems. Many efforts have been devoted to develop efficient algorithms for solving the optimization problem in nuclear norm regularization. Solving it for large-scale matrix variables, however, is still a challenging task since the complexity grows fast with the size of matrix variable. In this work, we propose a novel method called safe subspace screening (SSS), to improve the efficiency of the solver for nuclear norm regularized least squares problems. Motivated by the fact that the low rank solution can be represented by a few subspaces, the proposed method accurately discards a predominant percentage of inactive subspaces prior to solving the problem to reduce problem size. Consequently, a much smaller problem is required to solve, making it more efficient than optimizing the original problem. The proposed SSS is safe, in that its solution is identical to the solution from the solver. In addition, the proposed SSS can be used together with any existing nuclear norm solver since it is independent of the solver. Extensive results on several synthetic and real data sets show that the proposed SSS is very effective in inactive subspace screening.

Cite this Paper


BibTeX
@InProceedings{pmlr-v37-zhoua15, title = {Safe Subspace Screening for Nuclear Norm Regularized Least Squares Problems}, author = {Zhou, Qiang and Zhao, Qi}, booktitle = {Proceedings of the 32nd International Conference on Machine Learning}, pages = {1103--1112}, year = {2015}, editor = {Bach, Francis and Blei, David}, volume = {37}, series = {Proceedings of Machine Learning Research}, address = {Lille, France}, month = {07--09 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v37/zhoua15.pdf}, url = {https://proceedings.mlr.press/v37/zhoua15.html}, abstract = {Nuclear norm regularization has been shown very promising for pursing a low rank matrix solution in various machine learning problems. Many efforts have been devoted to develop efficient algorithms for solving the optimization problem in nuclear norm regularization. Solving it for large-scale matrix variables, however, is still a challenging task since the complexity grows fast with the size of matrix variable. In this work, we propose a novel method called safe subspace screening (SSS), to improve the efficiency of the solver for nuclear norm regularized least squares problems. Motivated by the fact that the low rank solution can be represented by a few subspaces, the proposed method accurately discards a predominant percentage of inactive subspaces prior to solving the problem to reduce problem size. Consequently, a much smaller problem is required to solve, making it more efficient than optimizing the original problem. The proposed SSS is safe, in that its solution is identical to the solution from the solver. In addition, the proposed SSS can be used together with any existing nuclear norm solver since it is independent of the solver. Extensive results on several synthetic and real data sets show that the proposed SSS is very effective in inactive subspace screening.} }
Endnote
%0 Conference Paper %T Safe Subspace Screening for Nuclear Norm Regularized Least Squares Problems %A Qiang Zhou %A Qi Zhao %B Proceedings of the 32nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2015 %E Francis Bach %E David Blei %F pmlr-v37-zhoua15 %I PMLR %P 1103--1112 %U https://proceedings.mlr.press/v37/zhoua15.html %V 37 %X Nuclear norm regularization has been shown very promising for pursing a low rank matrix solution in various machine learning problems. Many efforts have been devoted to develop efficient algorithms for solving the optimization problem in nuclear norm regularization. Solving it for large-scale matrix variables, however, is still a challenging task since the complexity grows fast with the size of matrix variable. In this work, we propose a novel method called safe subspace screening (SSS), to improve the efficiency of the solver for nuclear norm regularized least squares problems. Motivated by the fact that the low rank solution can be represented by a few subspaces, the proposed method accurately discards a predominant percentage of inactive subspaces prior to solving the problem to reduce problem size. Consequently, a much smaller problem is required to solve, making it more efficient than optimizing the original problem. The proposed SSS is safe, in that its solution is identical to the solution from the solver. In addition, the proposed SSS can be used together with any existing nuclear norm solver since it is independent of the solver. Extensive results on several synthetic and real data sets show that the proposed SSS is very effective in inactive subspace screening.
RIS
TY - CPAPER TI - Safe Subspace Screening for Nuclear Norm Regularized Least Squares Problems AU - Qiang Zhou AU - Qi Zhao BT - Proceedings of the 32nd International Conference on Machine Learning DA - 2015/06/01 ED - Francis Bach ED - David Blei ID - pmlr-v37-zhoua15 PB - PMLR DP - Proceedings of Machine Learning Research VL - 37 SP - 1103 EP - 1112 L1 - http://proceedings.mlr.press/v37/zhoua15.pdf UR - https://proceedings.mlr.press/v37/zhoua15.html AB - Nuclear norm regularization has been shown very promising for pursing a low rank matrix solution in various machine learning problems. Many efforts have been devoted to develop efficient algorithms for solving the optimization problem in nuclear norm regularization. Solving it for large-scale matrix variables, however, is still a challenging task since the complexity grows fast with the size of matrix variable. In this work, we propose a novel method called safe subspace screening (SSS), to improve the efficiency of the solver for nuclear norm regularized least squares problems. Motivated by the fact that the low rank solution can be represented by a few subspaces, the proposed method accurately discards a predominant percentage of inactive subspaces prior to solving the problem to reduce problem size. Consequently, a much smaller problem is required to solve, making it more efficient than optimizing the original problem. The proposed SSS is safe, in that its solution is identical to the solution from the solver. In addition, the proposed SSS can be used together with any existing nuclear norm solver since it is independent of the solver. Extensive results on several synthetic and real data sets show that the proposed SSS is very effective in inactive subspace screening. ER -
APA
Zhou, Q. & Zhao, Q.. (2015). Safe Subspace Screening for Nuclear Norm Regularized Least Squares Problems. Proceedings of the 32nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 37:1103-1112 Available from https://proceedings.mlr.press/v37/zhoua15.html.

Related Material