Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

DPxFin: Adaptive Differential Privacy for Anti-Money Laundering Detection via Reputation-Weighted Federated Learning

Renuga Kanagavelu, Manjil Nepal, Ning Peiyan, Cai Kangning, Xu Jiming, Fei Gao, Yong Liu, Goh Siow Mong Rick, Qingsong Wei

Abstract

In the modern financial system, combating money laundering is a critical challenge complicated by data privacy concerns and increasingly complex fraud transaction patterns. Although federated learning (FL) is a promising problem-solving approach as it allows institutions to train their models without sharing their data, it has the drawback of being prone to privacy leakage, specifically in tabular data forms like financial data. To address this, we propose DPxFin, a novel federated framework that integrates reputation-guided adaptive differential privacy. Our approach computes client reputation by evaluating the alignment between locally trained models and the global model. Based on this reputation, we dynamically assign differential privacy noise to client updates, enhancing privacy while maintaining overall model utility. Clients with higher reputations receive lower noise to amplify their trustworthy contributions, while low-reputation clients are allocated stronger noise to mitigate risk. We validate DPxFin on the Anti-Money Laundering (AML) dataset under both IID and non-IID settings using Multi Layer Perceptron (MLP). Experimental analysis established that our approach has a more desirable trade-off between accuracy and privacy than those of traditional FL and fixed-noise Differential Privacy (DP) baselines, where performance improvements were consistent, even though on a modest scale. Moreover, DPxFin does withstand tabular data leakage attacks, proving its effectiveness under real-world financial conditions.

DPxFin: Adaptive Differential Privacy for Anti-Money Laundering Detection via Reputation-Weighted Federated Learning

Abstract

In the modern financial system, combating money laundering is a critical challenge complicated by data privacy concerns and increasingly complex fraud transaction patterns. Although federated learning (FL) is a promising problem-solving approach as it allows institutions to train their models without sharing their data, it has the drawback of being prone to privacy leakage, specifically in tabular data forms like financial data. To address this, we propose DPxFin, a novel federated framework that integrates reputation-guided adaptive differential privacy. Our approach computes client reputation by evaluating the alignment between locally trained models and the global model. Based on this reputation, we dynamically assign differential privacy noise to client updates, enhancing privacy while maintaining overall model utility. Clients with higher reputations receive lower noise to amplify their trustworthy contributions, while low-reputation clients are allocated stronger noise to mitigate risk. We validate DPxFin on the Anti-Money Laundering (AML) dataset under both IID and non-IID settings using Multi Layer Perceptron (MLP). Experimental analysis established that our approach has a more desirable trade-off between accuracy and privacy than those of traditional FL and fixed-noise Differential Privacy (DP) baselines, where performance improvements were consistent, even though on a modest scale. Moreover, DPxFin does withstand tabular data leakage attacks, proving its effectiveness under real-world financial conditions.
Paper Structure (14 sections, 10 equations, 4 figures, 4 tables, 2 algorithms)

This paper contains 14 sections, 10 equations, 4 figures, 4 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Related Works
  3. Methodology of DPxFin
  4. Feature Engineering and Data Preparation
  5. Reputation-Based Adaptive Differential Privacy
  6. Server-Side Reputation Weighted Aggregation Phase
  7. Experimental Setup
  8. Dataset
  9. Computational Resources and Frameworks
  10. Evaluation of Performance Metrics
  11. Result and Analysis
  12. Analysis Under Different Privacy Budget
  13. Tabular Attack
  14. Conclusions

Figures (4)

  • Figure 1: Framework Overview: (1) Server shares the global model and reputation factors (2) Clients perform local training using DP-SGD with adaptive noise (3) Noisy updates are returned (4) Server measures contributions (5) Calculates the reputation scores and updates reputation factors (6) Performs reputation weighted aggregation (7) Repeat the cycle.
  • Figure 2: Dataset distribution among clients under a non-IID setting.
  • Figure 3: Accuracy comparison under IID and Non-IID distributions
  • Figure 4: Performance Across Different Federated Settings