English
 
Help Privacy Policy Disclaimer
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT

Released

Conference Paper

A scalable trust-region algorithm with application to mixed-norm regression

MPS-Authors
/persons/resource/persons76142

Sra,  S
Department Empirical Inference, Max Planck Institute for Biological Cybernetics, Max Planck Society;
Max Planck Institute for Biological Cybernetics, Max Planck Society;

Fulltext (restricted access)
There are currently no full texts shared for your IP range.
Fulltext (public)

ICML2010-Kim_6519[0].pdf
(Any fulltext), 333KB

Supplementary Material (public)
There is no public supplementary material available
Citation

Kim, D., Sra, S., & Dhillon, I. (2010). A scalable trust-region algorithm with application to mixed-norm regression. In J. Fürnkranz, & T. Joachims (Eds.), 27th International Conference on Machine Learning (ICML 2010) (pp. 519-526). Madison, WI, USA: Omnipress.


Cite as: https://hdl.handle.net/11858/00-001M-0000-0013-BF94-D
Abstract
We present a new algorithm for minimizing a convex loss-function subject to regularization. Our framework applies to numerous problems in machine learning and statistics; notably, for sparsity-promoting regularizers such as ℓ1 or ℓ1, ∞ norms, it enables efficient computation of sparse solutions. Our approach is based on the trust-region framework with nonsmooth objectives, which allows us to build on known results to provide convergence analysis. We avoid the computational overheads associated with the conventional Hessian approximation used by trust-region methods by instead using a simple separable quadratic approximation. This approximation also enables use of proximity operators for tackling nonsmooth regularizers. We illustrate the versatility of our resulting algorithm by specializing it to three mixed-norm regression problems: group lasso [36], group logistic regression [21], and multi-task lasso [19]. We experiment with both synthetic and real-world large-scale data—our method is seen to be competitive, robust, and scalable.