Please send questions or comments about PROC SCADLS to
This email address is being protected from spam bots, you need Javascript enabled to view it
This email address is being protected from spam bots, you need Javascript enabled to view it .
Variable selection is a classic and still very important problem in applied statistics. Methodology Center researcher Dr.
Runze Li and his collaborators have been studying penalized-likelihood approaches to automatic variable selection for predictive modeling when the number of potential predictors is very large. This involves adjusting the likelihood function using a penalty which rewards parsimony. One such penalty function, which has some favorable theoretical properties, is the Smoothly Clipped Absolute Deviation (SCAD) penalty (see Fan & Li 2001).
The Methodology Center has recently released a
SAS procedure, PROC SCADLS, which does linear regression using penalized least squares with the SCAD penalty. This approach automatically selects and estimates a model at the same time. In 2008 we hope to release PROC SCADGLIM, which will allow SCAD-penalized generalized linear models.
Fan, J., & Li, R. (2001). Variable selection via nonconcave penalized likelihood and its oracle properties.
Journal of the American Statistical Association, 96(456), 1348-1360. (For a reprint, please email
This email address is being protected from spam bots, you need Javascript enabled to view it
.)
The SCAD Penalty
