From a confusion matrix, how do you compute precision and recall?

• A. Precision = TP/(TP+FN); Recall = TP/(TP+FP).
• B. Precision = (TP+FP)/TP; Recall = (TP+FN)/TP.
• C. Precision = TP/(TP+FP); Recall = TP/(TP+FN).
• D. Precision = TP/(FN+TN); Recall = TP/(FP+TN).

Answer: (C)

In a confusion matrix, precision and recall come from comparing what the model predicted as positive with what was actually positive. Precision answers the question: of all the instances labeled positive by the model, how many are truly positive? This is computed as TP divided by (TP plus FP). Recall answers: of all the truly positive instances, how many did the model identify as positive? This is TP divided by (TP plus FN). FP lowers precision because it adds to the predicted positives without being true positives, while FN lowers recall because it adds to the actual positives that were missed. The expressions given match these definitions, with precision = TP/(TP+FP) and recall = TP/(TP+FN).