Bài kiểm tra có liên kết hướng dẫn · Cứng Cấp độ
Bài kiểm tra phân loại cảnh báo AML và giảm dương tính giả
Ưu tiên cảnh báo AML mà không xóa hoạt động đáng ngờ, sử dụng đánh giá của con người, dán nhãn cẩn thận, theo dõi kiểm tra và các biện pháp chất lượng.
Câu hỏi 1 của 8
How does AML alert triage support an investigation team?
Xem trước mọi câu hỏi8 câu hỏi
Câu trả lời và giải thích xuất hiện khi bạn làm bài kiểm tra.
- How does AML alert triage support an investigation team?
- Decide which account is guilty without analyst review
- Automatically close every low score
- Organize and prioritize alerts for appropriate investigation
- Replace customer due diligence
- Why is a non-filed alert not automatically a confirmed false positive?
- Every alert is required to become a SAR
- Disposition labels are selective and may not reflect true underlying activity
- A model has directly observed criminal intent
- A validated alert cannot be a false positive
- What risk comes from reducing alerts too aggressively?
- Investigation time always increases
- Customer data becomes less private
- The model stops using transaction data
- Suspicious patterns or new typologies may be missed
- How should a low model score be used in a risk workflow?
- One input to review under approved policy, not automatic proof of no risk
- A guarantee to close the case in every institution
- A public explanation of monitoring thresholds
- A reason to delete supporting records
- Which information helps investigators evaluate a prioritized alert?
- Only the final model score
- Only a customer name
- Transaction timeline, source records, linked entities and uncertainty
- A generated narrative without references
- Why sample cases that a triage model deprioritizes or closes?
- To increase the model's score
- To test for missed suspicious activity and policy gaps
- To remove audit requirements
- To confirm every low score is safe
- What should be logged if automated closure is allowed by approved policy?
- Only the processing time
- The private rule thresholds on a public page
- Nothing after the case is closed
- Policy basis, model version, evidence and decision rationale
- Which metric is more meaningful than simply counting fewer alerts?
- Review quality, missed-case risk, disposition time and customer impact
- The number of model parameters alone
- GPU utilization alone
- Only the total number of closed cases