Detecting Money Laundering with Node2Vec: An Innovative Approach Using Machine Learning

Analytical presentation by MSc. André Monteiro (AI Manager at Banco do Brasil, Master in BI) — summary of Caglayan, M. G., & Bahtiyar, S. (2022), “Money Laundering Detection with Node2Vec”, Gazi University Journal of Science.

What you will find:

• The problem: money laundering is 2–5% of global GDP ($800B–$2T); traditional detection mechanisms are losing effectiveness
• Method: Node2Vec graph representation + Naive Bayes and K-NN classifiers
• Class imbalance: SMOTE and ADASYN oversampling for minority (suspicious) samples
• Findings: graph representation + oversampling significantly improves detection; higher dimensions → better predictions (ROC/AUC)
• Implication: graph ML as a path to transform financial fraud detection

Format: PDF, 3 pages, English. Immediate download after purchase.