Sparse Features for PCA-Like Linear Regression

Principal Components Analysis (PCA) is often used as a feature extraction procedure.