Starlit: Privacy-Preserving Federated Learning to Enhance Financial Fraud Detection

TL;DR

Starlit is a scalable, privacy-preserving federated learning framework designed to improve financial fraud detection by addressing security, scalability, and dropout resistance issues present in prior solutions.

Contribution

This paper introduces Starlit, a novel federated learning mechanism that overcomes key limitations of existing methods, including security proofs, scalability, and client dropout handling.

Findings

Starlit demonstrates high scalability and efficiency in synthetic financial data tests.

The framework achieves accurate fraud detection without compromising privacy.

Starlit's design effectively handles client dropouts and security concerns.

Abstract

Federated Learning (FL) is a data-minimization approach enabling collaborative model training across diverse clients with local data, avoiding direct data exchange. However, state-of-the-art FL solutions to identify fraudulent financial transactions exhibit a subset of the following limitations. They (1) lack a formal security definition and proof, (2) assume prior freezing of suspicious customers' accounts by financial institutions (limiting the solutions' adoption), (3) scale poorly, involving either $O(n^2)$ computationally expensive modular exponentiation (where $n$ is the total number of financial institutions) or highly inefficient fully homomorphic encryption, (4) assume the parties have already completed the identity alignment phase, hence excluding it from the implementation, performance evaluation, and security analysis, and (5) struggle to resist clients' dropouts. This work introduces Starlit, a novel scalable privacy-preserving FL mechanism that overcomes these limitations. It has various applications, such as enhancing financial fraud detection, mitigating terrorism, and enhancing digital health. We implemented Starlit and conducted a thorough performance analysis using synthetic data from a key player in global financial transactions. The evaluation indicates Starlit's scalability, efficiency, and accuracy.