Credit-Card Fraud Detection: Decisions and System Design

Context

You are designing a real-time decisioning system for card transactions with strict latency constraints. At authorization time, only a subset of features is available; post-transaction outcomes (e.g., chargebacks) arrive weeks later.

For the accept/decline exercise, assume the following two minimal transaction snippets are available at decision time:

• Transaction A
  • Channel: e-commerce (card-not-present)
  • Amount: $4,200
  • Cardholder home country: UK
  • Merchant country: US (electronics)
  • Local time at cardholder: 03:17
  • Device fingerprint: new to platform
  • IP geolocation: NG (Nigeria), VPN likely
  • Velocity: 5 auth attempts in last 10 minutes on this card across different merchants
• Transaction B
  • Channel: card-present (EMV chip + PIN)
  • Amount: $18.75
  • Cardholder home country: UK
  • Merchant country: UK (coffee shop)
  • Local time at cardholder: 12:41
  • Device/terminal: known merchant terminal, low dispute history
  • Velocity: consistent with past user pattern (daily coffee purchases)

Tasks

1. Decide whether to accept or decline each transaction and justify your reasoning. If a conditional action (e.g., step-up authentication) is preferable, state it.
2. Outline a full fraud-detection strategy for card transactions: data required, feature engineering, real-time rules, and model-based approaches (including system/latency considerations).
3. Explain how unsupervised learning can be applied to detect fraudulent transactions and list suitable algorithms.
4. Detail appropriate evaluation metrics for fraud models, including how to assess unsupervised methods without labeled data.