A full-stack application designed to detect fraudulent transactions in real-time using machine learning.
This project integrates a Java Spring Boot backend, React frontend, and a Python Flask ML microservice with a MySQL database.
- 🔐 Real-time fraud prediction via REST API
- 🧠 ML microservice using Random Forest Classifier for high accuracy
- 🗂️ Modular architecture separating frontend, backend, and ML components
- 📊 Transaction data dashboard for visual analysis (in progress)
- 💾 Secure MySQL database integration for transaction storage
- 💬 API communication between Java backend and Python ML model
| Category | Technologies |
|---|---|
| Backend | Java, Spring Boot, REST APIs |
| Frontend | React.js |
| Machine Learning | Python, Flask, Scikit-learn |
| Database | MySQL |
| Tools | Git, GitHub, Postman, MySQL Workbench |
- Dataset: Financial transactions with labeled fraud cases
- Model: Random Forest Classifier trained for fraud detection
- Handles imbalanced data using techniques like SMOTE or class weighting
- Served via Flask API to provide real-time predictions
FraudDetection/
├── backend/ # Spring Boot backend source code
├── frontend/ # React frontend application
├── ml-model/ # Python Flask ML microservice
└── README.md # This file
- Java 11+ and Maven/Gradle for backend
- Node.js and npm for frontend
- Python 3.x with required packages (
flask,scikit-learn, etc.) - MySQL server
- Clone the repo
git clone https://github.com/Cherryga/FraudDetection.git
cd FraudDetection- Start ML microservice
cd ml-model
pip install -r requirements.txt
python app.py- Start backend
cd ../backend
# Run with IDE or:
./mvnw spring-boot:run- Start frontend
cd ../frontend
npm install
npm start- Backend & ML microservice integrated ✅
- Frontend UI functional with prediction feature ✅
- Planned: Data visualization dashboard, explainability tools (SHAP/LIME) 🔜
“Code is the silent guardian of secure transactions.” 🛡️