Computer scienceData scienceMachine learningClassificationClassification metrics

ROC and AUC

Probabilities

Report a typo

Suppose you've got the following list of probabilities from your model:

[0.75, 0.82, 0.32, 0.64, 0.81, 0.21, 0.47, 0.2, 0.19, 0.11]

You'd like to build a ROC curve now. For starters, you took a random decision threshold, let's say $0.74$ . Based on the decision threshold, assign the probabilities to the classes (1 or 0). The answer should be the count of $1$ 's in the obtained list.

Enter a number

___

Create a free account to access the full topic