Opening the paper…
Figure from the original question paper Figure from the original question paper Consider a binary classification dataset with labeled as 98% negative samples and 2% positive samples. A model is trained on this data, which of the following evaluation metrics are suitable for measuring effectiveness of this model: