Exercise

Confusion matrix calculations

Your classifier on the credit data achieved the following statistics: 163 true positives, 15 false positives, 48 false negatives, and 24 true negatives. These numbers are preloaded in the console environment for you as tp, fp, fn and tn respectively. The following statements involve two metrics: accuracy, given by the proportion of examples classified correctly, and recall, which is the proportion of truly positive examples that were classified as positive. Which of the statements is true?

Instructions

50 XP

Possible answers

The recall of the classifier is approximately 0.916

The recall of the classifier is approximately 0.63

The accuracy of the classifier is approximately 0.27

The recall of the classifier is approximately 0.77