Financial security / Machine learning
Financial Fraud Detection Project
Evaluating machine learning approaches to detect fraud in mobile money transactions.
About the project
This project uses MoMTSim to simulate mobile money transactions and known fraud scenarios in Sub-Saharan Africa. It compares six machine learning classifiers on the resulting synthetic datasets, evaluating how well they identify fraudulent transactions in imbalanced data. The research considers both detection performance and computational demands to inform model selection for mobile money platforms.
Project demonstration
Related paper
Financial Fraud Detection Using Rich Mobile Money Transaction Datasets
Read paper