Question: You and your colleague are working on a classification problem. You have approximately 6000 observations and about 15 highly correlated predictors. You both agree on
You and your colleague are working on a classification problem. You have approximately 6000 observations and about 15 highly correlated predictors. You both agree on using a penalized logistic regression method. You also suggest cross-validation to determine the model's tuning (penalty) parameter, but your colleague does not think that cross-validation is a good idea. What would be her argument against cross-validation?
Step by Step Solution
3.42 Rating (158 Votes )
There are 3 Steps involved in it
Your colleagues argument against crossvalidation for determining the models tuning parameter in the context of penalized logistic regression might be ... View full answer
Get step-by-step solutions from verified subject matter experts
