I have a dataset with around 30 independent variables and would like to construct a GLM to explore the relationship between them and the dependent variable. I am aware that the method I was taught ...
Is it always a good idea to train with the full dataset after k-fold cross-validation? Or is it better instead to stick with one of the models learned in one of the cross-validation splits for ...
[Edit] In light of the feedback I have received, I have added some more information below, to help clarify what I am seeking to 'model'. Additionally, I have reformualted the problem from a "dog ...