Training neural network classifiers for medical decision making: The effects of imbalanced datasets on classification performance