Bundle Methods for Regularized Risk Minimization

Simple item page
dc.contributor.authorTeo, Choon-Hui
dc.contributor.authorVishwanathan, S.V.N.
dc.contributor.authorSmola, Alexander
dc.contributor.authorQuoc, V, Le
dc.date.accessioned2015-12-07T22:21:47Z
dc.date.issued2010
dc.date.updated2016-02-24T11:13:48Z
dc.description.abstractA wide variety of machine learning problems can be described as minimizing a regularized risk functional, with different algorithms using different notions of risk and different regularizers. Examples include linear Support Vector Machines (SVMs), Gaussian Processes, Logistic Regression, Conditional Random Fields (CRFs), and Lasso amongst others. This paper describes the theory and implementation of a scalable and modular convex solver which solves all these estimation problems. It can be parallelized on a cluster of workstations, allows for data-locality, and can deal with regularizers such as L1 and L2 penalties. In addition to the unified framework we present tight convergence bounds, which show that our algorithm converges in O(1/ε) steps to e precision for general convex problems and in O(log(1/ε)) steps for continuously differentiable problems. We demonstrate the performance of our general purpose solver on a variety of publicly available data sets.
dc.identifier.issn1532-4435
dc.identifier.urihttp://hdl.handle.net/1885/20201
dc.publisherMIT Press
dc.sourceJournal of Machine Learning Research
dc.source.urihttp://jmlr.csail.mit.edu/papers/v11/
dc.subjectKeywords: Bundle methods; Cutting plane method; Cutting plane methods; Parallel optimization; Risk minimization; Subgradient methods; Convergence of numerical methods; Learning algorithms; Support vector machines; Optimization Bundle methods; Cutting plane method; Optimization; Parallel optimization; Regularized risk minimization; Subgradient methods
dc.titleBundle Methods for Regularized Risk Minimization
dc.typeJournal article
local.bibliographicCitation.lastpage365
local.bibliographicCitation.startpage311
local.contributor.affiliationTeo, Choon-Hui, College of Engineering and Computer Science, ANU
local.contributor.affiliationVishwanathan, S.V.N., Purdue University
local.contributor.affiliationSmola, Alexander, Yahoo! Research
local.contributor.affiliationQuoc, V, Le, Stanford University
local.contributor.authoremailrepository.admin@anu.edu.au
local.contributor.authoruidTeo, Choon-Hui, u4259204
local.description.embargo2037-12-31
local.description.notesImported from ARIES
local.identifier.absfor080109 - Pattern Recognition and Data Mining
local.identifier.absseo890205 - Information Processing Services (incl. Data Entry and Capture)
local.identifier.ariespublicationu4607519xPUB11
local.identifier.citationvolume11
local.identifier.scopusID2-s2.0-76749161402
local.identifier.uidSubmittedByu4607519
local.type.statusPublished Version

Downloads

Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
01_Teo_Bundle_Methods_for_Regularized_2010.pdf
Size:
1.75 MB
Format:
Adobe Portable Document Format
Download

Collections

ANU Research Publications

DSpace software copyright © 2002-2024 LYRASIS