Sep-12-2022, 12:42 PM
Hi all.
There are tons of examples about precision and recall online.
But I have difficulty explaining by connecting both metrics to the use case.
I would like to therefore seek you help if my following interpretation of both this metrics is correct? Else, do correct me. And what other info can we draw from this metrics?
"The classifier is able to correctly detect 76% of the negatives, however, it is only about 38% of the time correct.
"The classifier however, is able to correctly detect about only 31% of the positives, however, it is about 70% of the time correct."
There are tons of examples about precision and recall online.
But I have difficulty explaining by connecting both metrics to the use case.
I would like to therefore seek you help if my following interpretation of both this metrics is correct? Else, do correct me. And what other info can we draw from this metrics?
"The classifier is able to correctly detect 76% of the negatives, however, it is only about 38% of the time correct.
"The classifier however, is able to correctly detect about only 31% of the positives, however, it is about 70% of the time correct."
Output: precision recall f1-score
-ve (minority class) 0.376995 0.758358 0.503627
+ve (majority class) 0.698347 0.308619 0.428065