Application of Deep Learning and Explainable AI in Money Laundering Detection

Analytical presentation by MSc. André Luiz Monteiro (AI Manager at Banco do Brasil, Master in BI) — critical review of deep learning (DL) and explainable AI (XAI) techniques in anti-money laundering, based on Kute, D. V. et al. (2022).

What you will find:

• The false-positive problem: AML systems reach up to 98% false positives — why accuracy alone is not enough
• Systematic review: 77 articles classified into six categories, from anomaly detection to behavioral modeling and link analysis
• The interpretability gap: 51% of ML methods used in AML are uninterpretable; black-box vs white-box trade-offs
• XAI as the path forward: integrating DL with LIME and SHAP to generate interpretable explanations and increase trust in AML systems
• Data challenges: lack of recent labeled real data — and how deep learning on graphs and NLP can help

Format: PDF, 13 pages, English. Immediate download after purchase.