Custom scoring in cross-validation
The scoring functions used in our models are often baked in (such as using cross-entropy in Logistic Regression). We do get some choices when cross-validating, however. For example, we can pick the regularization parameter by using the ROC area...