Your question is Anomaly Detection for Fees. 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 given a dataset containing irregular trading settlement fees. How would you identify anomalies and clean the data for a cost-benefit write-up?