Your question is Imbalanced Dataset Metrics. 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).
What metrics do you use to evaluate models on imbalanced datasets?
Explain which metrics you would prioritize, why accuracy can be misleading, and how you would use threshold analysis and error inspection. Address how the choice changes when false positives and false negatives have different costs. No dataset, model, or metric values are provided, so state assumptions clearly.