Manifold Regularization for SIR with Rate Root-n Convergence

Bian, W; Tao, D

Manifold Regularization for SIR with Rate Root-n Convergence

Bian, W Tao, D

Permalink

Publisher:: Curran Associates, Inc
Publication Type:: Conference Proceeding
Citation:: Proceedings of the 2009 Conference ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 22, 2009, pp. 1 - 9
Issue Date:: 2009-01

Closed Access

	Filename	Description	Size
	2011001833OK.pdf		273.22 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	Bian, W	en_US
dc.contributor.author	Tao, D https://orcid.org/0000-0001-7225-5449	en_US
dc.contributor.editor	Bengio, Y	en_US
dc.contributor.editor	Schuurmans, D	en_US
dc.contributor.editor	Lafferty, J	en_US
dc.contributor.editor	Williams, C	en_US
dc.contributor.editor	Culotta, A	en_US
dc.date	2009-12-07	en_US
dc.date.issued	2009-01	en_US
dc.identifier.citation	Proceedings of the 2009 Conference ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 22, 2009, pp. 1 - 9	en_US
dc.identifier.isbn	9781615679119	en_US
dc.identifier.uri	http://hdl.handle.net/10453/17680
dc.description.abstract	In this paper, we study the manifold regularization for the Sliced Inverse Regression (SIR). The manifold regularization improves the standard SIR in two aspects: 1) it encodes the local geometry for SIR and 2) it enables SIR to deal with transductive and semi-supervised learning problems. We prove that the proposed graph Laplacian based regularization is convergent at rate root-n. The projection directions of the regularized SIR are optimized by using a conjugate gradient method on the Grassmann manifold. Experimental results support our theory.	en_US
dc.publisher	Curran Associates, Inc	en_US
dc.relation.ispartof	Proceedings of the 2009 Conference ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 22	en_US
dc.relation.ispartof	Annual Conference on Neural Information Processing Systems	en_US
dc.title	Manifold Regularization for SIR with Rate Root-n Convergence	en_US
dc.type	Conference Proceeding
utslib.location	Vancouver	en_US
utslib.location.activity	Vancouver, British Columbia, Canada	en_US
utslib.for	080109 Pattern Recognition and Data Mining	en_US
dc.location.activity	Vancouver, British Columbia, Canada	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 Computer Science
pubs.organisational-group	/University of Technology Sydney/Strength - CAI - Centre for Artificial Intelligence
utslib.copyright.status	closed_access
pubs.consider-herdc	false	en_US
pubs.place-of-publication	Vancouver	en_US
pubs.start-date	2009-12-07	en_US

Abstract:

In this paper, we study the manifold regularization for the Sliced Inverse Regression (SIR). The manifold regularization improves the standard SIR in two aspects: 1) it encodes the local geometry for SIR and 2) it enables SIR to deal with transductive and semi-supervised learning problems. We prove that the proposed graph Laplacian based regularization is convergent at rate root-n. The projection directions of the regularized SIR are optimized by using a conjugate gradient method on the Grassmann manifold. Experimental results support our theory.

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

http://hdl.handle.net/10453/17680