Learning From Very-Few Labeled Examples with Soft Labels

Mu, Y; Xu, M; Yan, S

Learning From Very-Few Labeled Examples with Soft Labels

Mu, Y Xu, M

Yan, S

Permalink

Publisher:: IEEE Computer Society
Publication Type:: Conference Proceeding
Citation:: 2010 IEEE International Conference on Image Processing ICIP 2010 - Proceedings, 2010, pp. 3869 - 3872
Issue Date:: 2010-01

Closed Access

	Filename	Description	Size
	2009007894OK.pdf		940.44 kB	Adobe PDF	View/Open

Copyright Clearance Process

Recently Added
In Progress
Closed Access

This item is closed access and not available.

Full metadata record

Field	Value	Language
dc.contributor.author	Mu, Y	en_US
dc.contributor.author	Xu, M https://orcid.org/0000-0001-9581-8849	en_US
dc.contributor.author	Yan, S	en_US
dc.contributor.editor	Technical Committee	en_US
dc.date	2010-09-26	en_US
dc.date.issued	2010-01	en_US
dc.identifier.citation	2010 IEEE International Conference on Image Processing ICIP 2010 - Proceedings, 2010, pp. 3869 - 3872	en_US
dc.identifier.isbn	978-1-4244-7993-1	en_US
dc.identifier.uri	http://hdl.handle.net/10453/16273
dc.description.abstract	In this paper we propose Softboost, a novel Boosting al-gorithm which combines the merits of transductive and inductive learning approaches to attack the problem of learning from very few labeled training examples. In the transductive stage, soft labels of both the labeled and unlabeled samples are estimated based on a Markovian propagating procedure. While in the subsequent inductive stage, to efficiently handle out-of-sample data, we learn a weighted combination of simple rules in Boosting style, each of which maximizes confidence-weighted inter-class Kullback-Leibler (KL) divergence under current data distribution. Finally, experiments on toy dataset and USPS handwritten digits are presented to demonstrate its effectiveness.	en_US
dc.publisher	IEEE Computer Society	en_US
dc.relation.ispartof	2010 IEEE International Conference on Image Processing ICIP 2010 - Proceedings	en_US
dc.relation.ispartof	IEEE International Conference on Image Processing	en_US
dc.rights	© 2010 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.	en_US
dc.title	Learning From Very-Few Labeled Examples with Soft Labels	en_US
dc.type	Conference Proceeding
utslib.location	Hongkong	en_US
utslib.location.activity	Hongkong	en_US
utslib.for	080106 Image Processing	en_US
dc.location.activity	Hongkong	en_US
pubs.embargo.period	Not known	en_US
pubs.organisational-group	/University of Technology Sydney
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology/School of Electrical and Data Engineering
pubs.organisational-group	/University of Technology Sydney/Strength - GBDTC - Global Big Data Technologies
pubs.organisational-group	/University of Technology Sydney/Strength - INEXT - Innovation in IT Services and Applications
utslib.copyright.status	closed_access
pubs.consider-herdc	true	en_US
pubs.place-of-publication	Hongkong	en_US
pubs.start-date	2010-09-26	en_US

Abstract:

In this paper we propose Softboost, a novel Boosting al-gorithm which combines the merits of transductive and inductive learning approaches to attack the problem of learning from very few labeled training examples. In the transductive stage, soft labels of both the labeled and unlabeled samples are estimated based on a Markovian propagating procedure. While in the subsequent inductive stage, to efficiently handle out-of-sample data, we learn a weighted combination of simple rules in Boosting style, each of which maximizes confidence-weighted inter-class Kullback-Leibler (KL) divergence under current data distribution. Finally, experiments on toy dataset and USPS handwritten digits are presented to demonstrate its effectiveness.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/16273