README.md

metadata

title: Fraud-Detection-System
sdk: streamlit
emoji: 🚀
colorFrom: indigo
colorTo: purple

💳 Credit Card & Transaction Fraud Detection System

Overview

A sophisticated real-time fraud detection system using ensemble machine learning models and advanced analytics to protect financial transactions.

🌟 Key Features

• Real-time Fraud Detection: Instant transaction monitoring and analysis
• Multi-Model Ensemble: XGBoost, LightGBM, Random Forest, and Gradient Boosting
• Advanced Analytics Dashboard: Interactive visualizations and insights
• Role-Based Access Control: User, Analyst, Manager, and Admin roles
• Email Alert System: Automated notifications for suspicious activities
• Batch Processing: Handle multiple transactions simultaneously

🚀 Technical Stack

• Frontend: Streamlit
• Backend: Python
• ML Models:
  • XGBoost
  • LightGBM
  • Random Forest
  • Gradient Boosting
• Data Processing: Pandas, NumPy
• Visualization: Plotly, Matplotlib
• Security: SHA-256 encryption

📊 Features In Detail

1. Fraud Detection

• Real-time transaction scoring
• Risk level assessment (LOW, MEDIUM, HIGH, CRITICAL)
• Behavioral analysis
• Location-based risk assessment
• Velocity checks

2. User Interface

• Interactive dashboards
• Real-time monitoring
• Advanced visualization
• Risk score breakdown
• Transaction patterns analysis

3. Security Features

• Role-based access control
• Password encryption
• Session management
• Activity logging
• Audit trails

4. Alert System

• Email notifications
• Risk-based alerting
• Customizable thresholds
• Batch alert processing

🚀 Dataset

The dataset used for this project is the Credit Card Fraud Detection dataset from Kaggle.

📁 Project Structure

Credit-Card-Fraud-Detection/
├── 📁 data/                      # Data directory
│   ├── users.csv                 # User credentials and roles
│   ├── transactions.csv          # Dataset of credit card transactions
│   └── models/                   # ML model implementations
│       ├── xgb_model.json        # XGBoost model file
│       └── features.pkl         # Saved feature configurations
│
├── 📁 .streamlit/               # Streamlit configurations
│   ├── config.toml             # Streamlit settings
│   └── secrets.toml            # Secure credentials
│
├── 📄 app.py                    # Main Streamlit application
├── 📄 requirements.txt          # Project dependencies
├── 📄 .env                      # Environment variables
├── 📄 .gitignore               # Git ignore patterns
├── 📄 LICENSE                   # License information
└── 📄 README.md                # Project documentation

🛠️ Installation

1. Clone the repository:

git clone https://github.com/D3V-S4NJ4Y/Credit-Card-Fraud-Detection-System

2. Install required packages:

pip install -r requirements.txt

4. Run the application:

streamlit run app.py

👥 User Roles

Default--> Usersname: Password

• admin: admin
• manager: 123456
• Register New User with Correct Email id for Email Alerts

1. User

   • View basic transaction details
   • Submit transactions
   • Receive alerts

2. Manager

   • Advanced analysis
   • Team management
   • Performance monitoring

3. Admin

   • Full system access
   • User management
   • System configuration

📈 Analytics Features

Transaction Analysis

• Risk scoring
• Behavioral patterns
• Anomaly detection
• Historical analysis

Visualization

• 3D risk analysis
• Decision flow diagrams
• Risk component breakdown
• Performance metrics

Reporting

• Real-time dashboards
• Batch analysis reports
• Custom filters
• Export capabilities

⚙️ Configuration

Email Setup

EMAIL_CONFIG = {
    "smtp_user": "your-email@gmail.com",
    "smtp_password": "your-app-password",
    "smtp_server": "smtp.gmail.com",
    "smtp_port": 587
}

Risk Thresholds

RISK_THRESHOLDS = {
    'CRITICAL': 0.9,
    'HIGH': 0.7,
    'MEDIUM': 0.5,
    'LOW': 0.2
}

🔒 Security Best Practices

1. Use environment variables for sensitive data
2. Regular password updates
3. Session timeout implementation
4. Activity logging
5. Data encryption

📝 Logs and Monitoring

• System health monitoring
• User activity logs
• Model performance tracking
• Error logging

📄 License

This project is licensed under the MIT License - see the LICENSE.md file for details.

👨‍💻 Author

Sanjay Kumar

• Email: sanjay.dev925@gmail.com
• GitHub: Your GitHub Profile

🙏 Acknowledgments

• Machine Learning libraries contributors
• Streamlit community
• Security framework developers

📞 Support

For support and queries:

• Create an issue in the repository
• Contact: sanjay.dev925@gmail.com

---

⭐ Star this repository if you find it helpful!