Lars Lasso
Lars Lasso, also known as Least Angle Regression with the L1 Penalty, is a machine-learning technique for feature selection and regularization. It builds upon Lasso regression and is effective for dealing with vast amounts of data. It works well even when there are more features than samples.