Students:
- Federico Romano Gargarella
Research Lead:
- Yasemin Ateş
This project analyzes Ethereum on-chain transaction data to build a practical fraud/risk triage pipeline like what a centralized crypto exchange (or an AML/compliance vendor) would use to decide which wallets/transactions to manually review.
The focus is on feature engineering from transaction behavior, training imbalanced-classification models, and turning model outputs into an operational “review-queue” policy under capacity constraints.