Group variable selection via SCAD-L2
Lingmin Zeng, Jun Xie
Abstract
Lingmin Zeng, Jun Xie
Abstract
We propose a penalized regression method SCAD-L2 using a penalty function called SCAD (smoothly clipped absolute deviation) combined with an L2 penalty. The new method inherits good features of SCAD, namely unbiasedness, continuity, and sparsity. In addition, it favours another important property that highly correlated variables are in or out a model together. SCAD-L2 derives its power by focusing on group variable selection. For data with dependent structures, where traditional variable selection methods are unstable, SCAD-L2 can select variable groups and preserve small prediction errors.
OpenAlex reports 52 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
We propose a penalized regression method SCAD-L2 using a penalty function called SCAD (smoothly clipped absolute deviation) combined with an L2 penalty. The new method inherits good features of SCAD, namely unbiasedness, continuity, and sparsity. In addition, it favours another important property that highly correlated variables are in or out a model together. SCAD-L2 derives its power by focusing on group variable selection. For data with dependent structures, where traditional variable selection methods are unstable, SCAD-L2 can select variable groups and preserve small prediction errors.
Key concepts: Scad, Mathematics, Feature selection, Variable (mathematics), Selection (genetic algorithm), Penalty method, Property (philosophy), Function (biology)