AML & Fraud Analytics Path · Mission 28 of 30Master

Rule precision scorecard with rank

Per-rule conditional aggregation of TP / FP / pending counts, precision %, and RANK by precision — the model-risk monthly review

Unlock with Pro→Back to AML & Fraud Analytics

The Brief

You'll practice

Conditional aggregationRANKNULLIF
← Previous
Mission 27
Next →
Mission 29
All 30 missions in AML & Fraud Analytics Path ↓
  1. 1. Meet the customer book
  2. 2. What channels do we see?
  3. 3. Active customers by state
  4. 4. Drop the reversals
  5. 5. Cash deposits over $10K
  6. 6. Sub-$10K cash deposits by customer
  7. 7. Daily deposit count per account
  8. 8. Dormant accounts that just woke up
  9. 9. Customers with no KYC refresh in 3 years
  10. 10. Wire vs cash — volume by channel group
  11. 11. Round-dollar wires — flag a pattern
  12. 12. Velocity — 5+ cash deposits in 24 hours
  13. 13. Structuring candidates — $9K–$9,999 over 14 days
  14. 14. Counterparty ring — same two customers, repeated
  15. 15. Device reuse across "separate" customers
  16. 16. Email reuse — fuzzy entity resolution
  17. 17. SOUNDEX-style name dedup across customer_ids
  18. 18. Alert precision — how good is our rule?
  19. 19. False positives by rule — tuning targets
  20. 20. Mule network — one receives, many send
  21. 21. Structuring ring capstone — everyone in the network
  22. 22. CTR / SAR filing candidate list
  23. 23. Sanctions-list fuzzy match scorecard
  24. 24. Rule-tuning backtest — precision quadrants
  25. 25. The investigator's dashboard
  26. 26. Multi-signal customer risk score
  27. 27. Cash-intensive customer profile
  28. 28. Rule precision scorecard with rank
  29. 29. Sanctions fuzzy match scorecard
  30. 30. Investigator capstone — case-load board