You are training a binary classifier where the positive class is rare, so standard training and evaluation can give misleading results.
How do you handle imbalanced datasets in machine learning?
