Introduction to Machine Learning (I2ML) | Chapter 04.10: ROC Curves

In this section, we explain the ROC curve and how to calculate it. In addition, we will present the AUC as a global performance measure that integrates over all possible thresholds.

Lecture video

Lecture slides

Code demo

You can run the code snippets in the demos on your local machine. The corresponding Rmd version of this demo can be found here. If you want to render the Rmd files to PDF, you need the accompanying style files.

Quiz

--- shuffle_questions: false --- ## Which statements are true? - [ ] If the proportion of positive to negative instances in the training data changes, the ROC curve will not change. - [x] If the proportion of positive to negative instances in the test data changes, the ROC curve will not change. - [x] Several evaluation metrics can be derived from a confusion matrix. - [x] The area under the ROC curve is called AUC. - [ ] A model with AUC=0 is the worst case. ## Which statements are true? - [ ] Using the prediction on the train data is the ordinary and correct way of calculating the ROC. - [x] The calculation of the ROC should be done on a test set. - [x] The AUC is not affected by the threshold, as opposed to the MCE.