Skip to content

utkrsh29/UniPay-FraudX

 
 

Repository files navigation

🚀 UniPay FraudX

AI-Powered Digital Payment Fraud Detection System


📌 Overview

UniPay FraudX is a machine learning-based web application designed to detect fraudulent and suspicious digital transactions in university and community ecosystems.

The system analyzes transaction patterns and classifies them as Normal or Suspicious, providing real-time insights through an interactive dashboard.


⚠️ Problem Statement

With the rapid growth of digital payments in academic and local communities, there is an increasing risk of fraudulent transactions.

Existing fraud detection systems are often complex, expensive, and not suitable for educational environments. There is a need for a simple, accessible, and intelligent system to identify suspicious activities and promote digital payment awareness.


🎯 Objectives

  • Analyze digital transaction patterns using Machine Learning
  • Identify and classify suspicious transactions
  • Provide clear and explainable results
  • Build an interactive and user-friendly dashboard
  • Promote awareness of digital payment risks
  • Enhance understanding of fraud detection concepts

💡 Proposed Solution

UniPay FraudX is a web-based application that uses machine learning to analyze transaction data and detect anomalies.

The system processes transaction inputs, applies a trained model, and classifies them into fraud or normal categories.

Users can visualize transaction insights, risk levels, and patterns through an interactive dashboard, making the system both educational and practical.


🛠️ Tech Stack

🔹 Backend

  • Python
  • Streamlit

🔹 Machine Learning

  • Scikit-learn (Logistic Regression)

🔹 Data Processing

  • Pandas
  • NumPy

🔹 Visualization

  • Plotly

🔹 Frontend

  • Streamlit UI
  • HTML / CSS

📊 Features

  • 📈 Interactive dashboard with charts
  • 🔍 Real-time fraud prediction
  • ⚠️ Risk level detection (Normal / Suspicious)
  • 🎨 Dark & Light theme support
  • 📊 Data visualization (Pie, Bar, Line charts)
  • 🧠 Machine learning-based classification
  • 🖥️ Clean and modern UI

🔄 System Workflow

  1. User inputs transaction details
  2. Data is preprocessed
  3. ML model analyzes the input
  4. Transaction is classified
  5. Result is displayed with insights
  6. Dashboard updates with visualization

📂 Dataset

  • AI-generated simulated dataset
  • Designed to mimic real-world digital transactions
  • Ensures privacy and safe academic usage

🌐 Deployment

The application is deployed using Streamlit Cloud and can be accessed online.

👉 Live Demo: https://minor-project-sem-4-2wdxbhegndunimrnis6i5r.streamlit.app/


📌 Future Improvements

  • Integration with real-time payment APIs
  • Advanced ML models (Random Forest, XGBoost)
  • Fraud probability scoring
  • User authentication system
  • Real dataset integration

👩‍💻 Author

Aabha Tomar
B.Tech CSE (Data Science)


🤝 Contributors

  • Aabha Tomar
  • Simran
  • Nishant Rajawat
  • Sakshi Rajawat
  • Chetna Sharma

📂 Project Structure

UniPay-FraudX/
├── dataset/              # Transaction datasets
├── models/               # Pre-trained ML models
├── backend_archive/      # (Archived) Older Flask backend files
├── templates/            # (Archived) Older HTML templates
├── static/               # (Archived) Older CSS files
├── app.py                # Main Streamlit Dashboard Application
├── requirements.txt      # Project dependencies
└── README.md             # Project documentation

⚙️ Installation

  1. Clone the repository:
git clone https://github.com/Aabhaaatomar/UniPay-FraudX
cd UniPay-FraudX
  1. Install the required dependencies:
pip install -r requirements.txt
  1. Run the application:
streamlit run app.py

⭐ Conclusion

UniPay FraudX demonstrates how machine learning can be applied to detect fraud in digital payments while maintaining simplicity and usability for academic environments.

About

UniPay FraudX : A Machine Learning-Based Digital Payment Fraud Detection System for University and Community Ecosystems

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages

  • Python 92.1%
  • CSS 5.4%
  • HTML 2.5%