Your question is Rare Failure Prediction Under Imbalance. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
You are training a supervised model to predict rare vehicle component failures from historical service and telemetry data. Positive examples are scarce, but missing a true failure is costly and too many false alerts will overwhelm operations.
How do you handle highly imbalanced datasets when training a model to predict rare vehicle component failures?