Question: Let's say that I have a classification problem that is very imbalanced: There are only a few yes cases, and a lot of no cases.
Let's say that I have a classification problem that is very imbalanced: There are only a few "yes" cases, and a lot of "no" cases.
In this situation, what is likely the worst error metric to use when evaluating potential classifiers on the training data?
A. Percent correct
B. Recall
C. F1
D. AUC
E. Any of the above would be fine
Step by Step Solution
There are 3 Steps involved in it
Get step-by-step solutions from verified subject matter experts
