Home News About Us Contact Contributors Disclaimer Privacy Policy Help FAQ

Home
Search
Quick Search
Advanced
Fulltext
Browse
Collections
Persons
My eDoc
Session History
Login
Name:
Password:
Documentation
Help
Support Wiki
Direct access to
document ID:


          Institute: MPI für Intelligente Systeme (ehemals Max-Planck-Institut für Metallforschung)     Collection: Abt. Schölkopf (Empirical Inference)     Display Documents



  history
ID: 596122.0, MPI für Intelligente Systeme (ehemals Max-Planck-Institut für Metallforschung) / Abt. Schölkopf (Empirical Inference)
Projected Newton-type methods in machine learning
Authors:Schmidt, M.; Kim, D.; Sra, S.
Place of Publication:Cambridge, MA, USA
Publisher:MIT Press
Date of Publication (YYYY-MM-DD):2011-12-01
Title of Book:Optimization for Machine Learning
Start Page:305
End Page:330
Physical Description:25
Full Name of Book-Editor(s):Sra, S.; Nowozin, S.; Wright, S. J.
Review Status:not specified
Audience:Not Specified
Intended Educational Use:No
Abstract / Description:We consider projected Newton-type methods for solving large-scale optimization problems arising in machine learning and related fields. We first introduce an algorithmic framework for projected Newton-type methods by reviewing a canonical projected (quasi-)Newton method. This method, while conceptually pleasing, has a high computation cost per iteration. Thus, we discuss two variants that are more scalable, namely, two-metric projection and inexact
projection methods. Finally, we show how to apply the Newton-type framework to handle non-smooth objectives. Examples are provided throughout the chapter to illustrate machine learning applications of our framework.
External Publication Status:published
Document Type:InBook
Communicated by:Heide Klooz
Affiliations:MPI für Intelligente Systeme/Abt. Schölkopf
Identifiers:URL:http://www.kyb.tuebingen.mpg.de//fileadmin/user_up...
LOCALID:6824
The scope and number of records on eDoc is subject to the collection policies defined by each institute - see "info" button in the collection browse view.