SeminarsGeneral Adaptive Sparse Principal Component Analysis
150
reads
The method of sparse principal component analysis (S-PCA) proposed by Zou, Hastie and Tibshirani (2006) is an attractive approach to obtain sparse loadings in principal component analysis (PCA). S-PCA was motivated by reformulating PCA as a least squares problem so that a lasso penalty on the loading coefficients can be applied. In this talk, we propose new estimates to improve S-PCA in terms of variable selection and efficiency. Our new formulation allows us to study many related sparse PCA estimators under a unified theoretical framework, and it leads to the method of general adaptive sparse principal component analysis (GAS-PCA) which is more attractive in many aspects. Numerical studies are conducted to compare the finite sample performance of various competing methods.
reads
Chenlei Leng
2008-12-05
13:30:00 - 15:00:00
405 , Mathematics Research Center Building (ori. New Math. Bldg.)
The method of sparse principal component analysis (S-PCA) proposed by Zou, Hastie and Tibshirani (2006) is an attractive approach to obtain sparse loadings in principal component analysis (PCA). S-PCA was motivated by reformulating PCA as a least squares problem so that a lasso penalty on the loading coefficients can be applied. In this talk, we propose new estimates to improve S-PCA in terms of variable selection and efficiency. Our new formulation allows us to study many related sparse PCA estimators under a unified theoretical framework, and it leads to the method of general adaptive sparse principal component analysis (GAS-PCA) which is more attractive in many aspects. Numerical studies are conducted to compare the finite sample performance of various competing methods.