Hello,
Regarding today's quizz, I've got a question about the computation of recall & precision metrics (on the reviews subject).
When computing such metrics, we want to have a dominant class to be determined as TRUE and one as FALSE. For example, if we had classification of dogs and cats, the precision metric for dogs isn't the same as the one for cats.
In today's quizz, no argument was made in the question description about the one to use as the dominant one. So I assumed, as you were talking about the negative only, we should directly compute the metrics over the negative class, and not reverse the whole thing (which is not that difficult btw, so it's not the problem here, but only the comprehension of the question itself).
I guess it's just my mistake now, but is it possible for the next quizzes/exam to clearly state which class we should compute the metrics on?
Also, if both questions have been well understood by everyone, it's just mb then.
Thank you!
