Spectral Bounds for Sparse PCA: Exact and Greedy Algorithms

Spectral Bounds for Sparse PCA: Exact and Greedy Algorithms (2006)

Cached

Download Links

[www.merl.com]
[www.merl.com]

Other Repositories/Bibliography

DBLP

by Baback Moghaddam , Yair Weiss , Shai Avidan

Venue:	Advances in Neural Information Processing Systems
Citations:	34 - 4 self

Versions

Version History

Metadata Version 2

User correction supplied by SystemCorrections

Datum	Value	Source
TITLE	Spectral Bounds for Sparse PCA: Exact and Greedy Algorithms	user correction - Legacy Corrections
AUTHOR NAME	Baback Moghaddam	SVM HeaderParse 0.2
AUTHOR ADDR	Cambridge MA, USA	SVM HeaderParse 0.2
AUTHOR NAME	Yair Weiss	SVM HeaderParse 0.2
AUTHOR AFFIL	Hebrew University	SVM HeaderParse 0.2
AUTHOR ADDR	Jerusalem, Israel	SVM HeaderParse 0.2
AUTHOR NAME	Shai Avidan	SVM HeaderParse 0.2
AUTHOR ADDR	Cambridge MA, USA	SVM HeaderParse 0.2
ABSTRACT	Sparse PCA seeks approximate sparse "eigenvectors" whose projections capture the maximal variance of data. As a cardinality-constrained and non-convex optimization problem, it is NP-hard and is encountered in a wide range of applied fields, from bio-informatics to finance. Recent progress has focused mainly on continuous approximation and convex relaxation of the hard cardinality constraint. In contrast, we consider an alternative discrete spectral formulation based on variational eigenvalue bounds and provide an effective greedy strategy as well as provably optimal solutions using branch-and-bound search. Moreover, the exact methodology used reveals a simple renormalization step that improves approximate solutions obtained by any continuous method. The resulting performance gain of discrete algorithms is demonstrated on real-world benchmark data and in extensive Monte Carlo evaluation trials.	user correction - Legacy Corrections
YEAR	2006	INFERENCE
VENUE	Advances in Neural Information Processing Systems	INFERENCE
VENUE TYPE	CONFERENCE	INFERENCE
PAGES	915--922	INFERENCE
VOLUME	18	INFERENCE
CITATIONS	13 found	ParsCit 1.0