Why is caret assuming BestModel = the one who minimize the CV-metric and not the minimum+1se like in LASSO?

222 views Asked by At

Usually, when approaching to LASSO, the best hyperparameter lambda is assumed to be lambda.1se (in the glmnet package), which is the lambda minimizing/maxining the CV-metric (usually AUC, Accuracy or Deviance) PLUS 1 standard deviation. lamnda.1se = lambda.min + sd(lambda.min)

In this way there is a stronger penalization and the model is less prone to overfitting.

However, in caret, I always find the assumption that the best model is the corresponding to the min/max of the CV-metric. Is it wise?

Is there a way to ask caret train function to select the model corresponding to lambda.1se?

Thank you.

0

There are 0 answers