Professional Machine Learning Engineer

Question

You recently deployed an ML model. Three months after deployment, you find that the model performs poorly for certain subgroups, potentially producing biased outcomes. You suspect this inequitable performance results from class imbalances in the training data, but you cannot gather additional data. What should you do?

Choose two

Explanation

Cost-sensitive training can assign a greater penalty to errors on the minority class, encouraging the model to improve performance for that class. Upsampling or reweighting existing training examples counteracts class imbalance without requiring new data, so both methods can reduce bias caused by underrepresented classes.

Professional Machine Learning Engineer

AstroTutor

Notes

Readiness

QuestionQ15

Community Discussion