Does deep learning model calibration improve performance in class-imbalanced medical image classification?

Re-assign community