High-dimensional Inference via Lipschitz Sparsity-Yielding Regularizers

Zheng Pan, Changshui Zhang
Proceedings of the Sixteenth International Conference on Artificial Intelligence and Statistics, PMLR 31:481-488, 2013.

Abstract

Non-convex regularizers are more and more applied to high-dimensional inference with sparsity prior knowledge. In general, the non-convex regularizer is superior to the convex ones in inference but it suffers the difficulties brought by local optimums and massive computation. A "good" regularizer should perform well in both inference and optimization. In this paper, we prove that some non-convex regularizers can be such "good" regularizers. They are a family of sparsity-yielding penalties with proper Lipschitz subgradients. These regularizers keep the superiority of non-convex regularizers in inference. Their estimation conditions based on sparse eigenvalues are weaker than the convex regularizers. Meanwhile, if properly tuned, they behave like convex regularizers since standard proximal methods guarantee to give stationary solutions. These stationary solutions, if sparse enough, are identical to the global solutions. If the solution sequence provided by proximal methods is along a sparse path, the convergence rate to the global optimum is on the order of 1/k where k is the number of iterations.

Cite this Paper


BibTeX
@InProceedings{pmlr-v31-pan13a, title = {High-dimensional Inference via Lipschitz Sparsity-Yielding Regularizers}, author = {Pan, Zheng and Zhang, Changshui}, booktitle = {Proceedings of the Sixteenth International Conference on Artificial Intelligence and Statistics}, pages = {481--488}, year = {2013}, editor = {Carvalho, Carlos M. and Ravikumar, Pradeep}, volume = {31}, series = {Proceedings of Machine Learning Research}, address = {Scottsdale, Arizona, USA}, month = {29 Apr--01 May}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v31/pan13a.pdf}, url = {https://proceedings.mlr.press/v31/pan13a.html}, abstract = {Non-convex regularizers are more and more applied to high-dimensional inference with sparsity prior knowledge. In general, the non-convex regularizer is superior to the convex ones in inference but it suffers the difficulties brought by local optimums and massive computation. A "good" regularizer should perform well in both inference and optimization. In this paper, we prove that some non-convex regularizers can be such "good" regularizers. They are a family of sparsity-yielding penalties with proper Lipschitz subgradients. These regularizers keep the superiority of non-convex regularizers in inference. Their estimation conditions based on sparse eigenvalues are weaker than the convex regularizers. Meanwhile, if properly tuned, they behave like convex regularizers since standard proximal methods guarantee to give stationary solutions. These stationary solutions, if sparse enough, are identical to the global solutions. If the solution sequence provided by proximal methods is along a sparse path, the convergence rate to the global optimum is on the order of 1/k where k is the number of iterations.} }
Endnote
%0 Conference Paper %T High-dimensional Inference via Lipschitz Sparsity-Yielding Regularizers %A Zheng Pan %A Changshui Zhang %B Proceedings of the Sixteenth International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2013 %E Carlos M. Carvalho %E Pradeep Ravikumar %F pmlr-v31-pan13a %I PMLR %P 481--488 %U https://proceedings.mlr.press/v31/pan13a.html %V 31 %X Non-convex regularizers are more and more applied to high-dimensional inference with sparsity prior knowledge. In general, the non-convex regularizer is superior to the convex ones in inference but it suffers the difficulties brought by local optimums and massive computation. A "good" regularizer should perform well in both inference and optimization. In this paper, we prove that some non-convex regularizers can be such "good" regularizers. They are a family of sparsity-yielding penalties with proper Lipschitz subgradients. These regularizers keep the superiority of non-convex regularizers in inference. Their estimation conditions based on sparse eigenvalues are weaker than the convex regularizers. Meanwhile, if properly tuned, they behave like convex regularizers since standard proximal methods guarantee to give stationary solutions. These stationary solutions, if sparse enough, are identical to the global solutions. If the solution sequence provided by proximal methods is along a sparse path, the convergence rate to the global optimum is on the order of 1/k where k is the number of iterations.
RIS
TY - CPAPER TI - High-dimensional Inference via Lipschitz Sparsity-Yielding Regularizers AU - Zheng Pan AU - Changshui Zhang BT - Proceedings of the Sixteenth International Conference on Artificial Intelligence and Statistics DA - 2013/04/29 ED - Carlos M. Carvalho ED - Pradeep Ravikumar ID - pmlr-v31-pan13a PB - PMLR DP - Proceedings of Machine Learning Research VL - 31 SP - 481 EP - 488 L1 - http://proceedings.mlr.press/v31/pan13a.pdf UR - https://proceedings.mlr.press/v31/pan13a.html AB - Non-convex regularizers are more and more applied to high-dimensional inference with sparsity prior knowledge. In general, the non-convex regularizer is superior to the convex ones in inference but it suffers the difficulties brought by local optimums and massive computation. A "good" regularizer should perform well in both inference and optimization. In this paper, we prove that some non-convex regularizers can be such "good" regularizers. They are a family of sparsity-yielding penalties with proper Lipschitz subgradients. These regularizers keep the superiority of non-convex regularizers in inference. Their estimation conditions based on sparse eigenvalues are weaker than the convex regularizers. Meanwhile, if properly tuned, they behave like convex regularizers since standard proximal methods guarantee to give stationary solutions. These stationary solutions, if sparse enough, are identical to the global solutions. If the solution sequence provided by proximal methods is along a sparse path, the convergence rate to the global optimum is on the order of 1/k where k is the number of iterations. ER -
APA
Pan, Z. & Zhang, C.. (2013). High-dimensional Inference via Lipschitz Sparsity-Yielding Regularizers. Proceedings of the Sixteenth International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 31:481-488 Available from https://proceedings.mlr.press/v31/pan13a.html.

Related Material