|
Activity Number:
|
318
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 4, 2009 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistical Learning and Data Mining
|
| Abstract - #304752 |
|
Title:
|
An Iterative Thresholding Approach for Sparse PCA
|
|
Author(s):
|
Zongming Ma*+
|
|
Companies:
|
Stanford University
|
|
Address:
|
Department of Statistics, 390 Serra Mall , Stanford, CA, 94305,
|
|
Keywords:
|
Principal component analysis ; sparsity ; iterative algorithm ; thresholding
|
|
Abstract:
|
Principal component analysis is a widely used statistical method. For reasons including interpretability and statistical properties, sparsity in the estimated principal components is desired in a number of situations. In this talk, we present an iterative thresholding approach for sparse PCA, which has a close connection to a Lasso-type formulation of this problem. In particular, we investigate various statistical properties of the proposed method in a high dimensional setting under some assumptions.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2009 program |
