2
$\begingroup$

Problem statement: Evaluate a binary classifier. There are 50 positive outcomes in the test data, and 100 observations. Using a 50% threshold, the classifier predicts 40 positive outcomes, of which 10 are incorrect. Calculate classifier’s Precision and Recall on the test sample?

My tried solution: I have calculated Precision as : P = TP / TP + FP => P = 40 / 40 + 10 => P = 0.8 And, same Recall. But, I am not sure its correct as F1 score options are not matching with mine.

So, I need to build a correct 2x2 Confusion matrix.

Please let me know for any query or information required to solve this question. Thanks

$\endgroup$

1 Answer 1

2
$\begingroup$

The number of true positives (TP) is 30 because you predicted 40 positives, however only 30 of them were correct. That would make your precision $$p=\frac{30}{40}=0.75$$

Similarly, for recall, it's $30/50$.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.