ML functional performance metrics · Question 44 of 60
A model classifies all 1,000 cases as positive; 50 are actually positive. Which values result?
3 answers are correct. Hard
Show answer
Answer
- B Accuracy 5%
- C Precision 5%
- E Recall 100%
Explanation
There are TP = 50, FP = 950, FN = 0 and TN = 0. Recall = 50 / 50 = 100%, because no positive case is missed. Precision = 50 / 1000 = 5%, and accuracy is also (50 + 0) / 1000 = 5%. A recall of 100% alone therefore says nothing about false alarms.