A hospital uses Azure Custom Vision to classify X-ray images as normal or abnormal. The model achieves 98% accuracy on the test set. However, during deployment, the model misclassifies many abnormal cases as normal, causing missed diagnoses. The hospital has a class imbalance where abnormal cases are only 5% of the data. What should the data scientist do first to address this?
Trap 1: Increase the number of training epochs.
More epochs may lead to overfitting and not address imbalance.
Trap 2: Add more normal X-ray images to the dataset.
Adding more normal images worsens the imbalance.
Trap 3: Switch to a different object detection algorithm.
The algorithm is not the root cause; imbalance is.
- A
Increase the number of training epochs.
Why it fails: More epochs may lead to overfitting and not address imbalance.
- B
Add more normal X-ray images to the dataset.
Why it fails: Adding more normal images worsens the imbalance.
- C
Switch to a different object detection algorithm.
Why it fails: The algorithm is not the root cause; imbalance is.
- D
Use oversampling or class-weight techniques to balance the training.
Balancing the dataset or adjusting loss weights improves minority class recall.