Sparse principal component analysis and iterative thresholding

Zongming Ma

Principal component analysis (PCA) is a classical dimension reduction method which projects data onto the principal subspace spanned by the leading eigenvectors of the covariance m…