Which formula correctly defines precision in binary classification?

Which formula correctly defines precision in binary classification?

• A. TP / (TP + FN)
• B. FN / (FN + TN)
• C. TP / (TP + FP)
• D. (TP + TN) / Total

Answer: (C)

Precision tells you how many of the items labeled positive by the model actually belong to the positive class. It is calculated as the number of true positives divided by all positives predicted by the model (true positives plus false positives). For example in a spam classifier, if you mark 100 emails as spam and 80 are truly spam while 20 are not, precision is 80/100 = 0.8. The other formulas mix different pieces: TP/(TP+FN) is recall, which measures how many of the actual positives you found, not how pure your positive predictions are. FN/(FN+TN) isn’t a standard metric and uses a denominator that doesn’t represent either the predicted positives or the actual totals. And (TP+TN)/Total is accuracy, the overall proportion of correct predictions across both classes. Therefore, precision is TP/(TP+FP).

Precision tells you how many of the items labeled positive by the model actually belong to the positive class. It is calculated as the number of true positives divided by all positives predicted by the model (true positives plus false positives). For example in a spam classifier, if you mark 100 emails as spam and 80 are truly spam while 20 are not, precision is 80/100 = 0.8. The other formulas mix different pieces: TP/(TP+FN) is recall, which measures how many of the actual positives you found, not how pure your positive predictions are. FN/(FN+TN) isn’t a standard metric and uses a denominator that doesn’t represent either the predicted positives or the actual totals. And (TP+TN)/Total is accuracy, the overall proportion of correct predictions across both classes. Therefore, precision is TP/(TP+FP).