Propozycja modyfikacji metody złagodzonego LASSO
The proposition of modification of the relaxed LASSO method

Summary/Abstract: Regularized linear regression (i.e. LASSO [Tibshirani 1996]) has reached a lot of interest as a feature selection tool. Meinshausen [2007] proposed a modified version of the LASSO by introducing a relaxation parameter which controls the variances of the pa-rameters, regardless of the feature elimination stage. This method is recommended in high dimensions, and for the high signal-to-noise ratio. The modification of the relaxed LASSO method is proposed in this paper. The simulations show that the new approach provides more stable results, and more effectively discards noisy variables.

• Details
• Contents