[R] Running GCV Optimization under Ridge Regression

Preetam Pal lordpreetam at gmail.com
Mon Sep 21 21:49:38 CEST 2015

Hi guys,

I am running Ridge regression on a dataset (predicted variable = y; GDP,
HPA and FX are regressors). I found that lm.ridge() can perform the ridge
regression given any value of lambda (i.e. the ridge parameter). However,
in order to choose the best results, I need to select the model output
corresponding to that lambda which optimizes some logically defined GCV
criteria. I thought there would be some in-built funcion in R Studio for
this, but could not find one. Also, I am not being able to write the
required code for this. Any help here will be appreciated. I have attached
the data in case it is required.


More information about the R-help mailing list