Chapter 04.09: ROC Basics

From the confusion matrix we can calculate a variety of ROC metrics. Among others, we will explain true positive rate, negative predictive value and the $F1$ measure.

Lecture video

Lecture slides

Quiz

--- shuffle_questions: false --- ## Which statements are true? - [x] Logistic regression minimizes the binomial loss. - [x] The Brier score is like the MSE, just with probabilities. - [ ] The log-loss punishes being very wrong less than the Brier score. - [ ] Accuracy and mean classification error are calculated using the predicted probabilities. - [x] The confusion matrix tabulates the true against predicted classes. - [ ] A misclassification error rate of 0.1% is always great.