A Machine Learning project that predicts whether a financial transaction is fraudulent or legitimate using supervised learning. The project covers the complete ML workflow—from data preprocessing and feature engineering to model deployment with an interactive Streamlit web application for real-time predictions.
Financial fraud is one of the biggest challenges in digital payment systems. This project leverages machine learning techniques to classify transactions as fraudulent or legitimate based on transaction-related features. The application is deployed using Streamlit, allowing users to make instant predictions through a simple web interface.
| Resource | Link |
|---|---|
| 🌐 Live App | Launch Fraud Detection Predictor |
| 💻 Source Code | GitHub Repository |
- Data preprocessing and cleaning
- Exploratory Data Analysis (EDA)
- Feature engineering
- Handling class imbalance
- One-Hot Encoding for categorical features
- Feature Scaling
- Machine Learning classification model
- Model evaluation using multiple metrics
- Interactive Streamlit web application
- Real-time fraud prediction
- Python
- Pandas
- NumPy
- Scikit-learn
- Matplotlib
- Seaborn
- Streamlit
- Joblib
Fraud-Detection-Predictor/
│
├── assets/
│ ├── home.png
│ ├── prediction.png
│ └── result.png
│
├── fraud_detection.py # Streamlit application
├── fraud_detection_pipeline.pkl # Trained ML pipeline
├── analysis_model.ipynb # Model development notebook
├── requirements.txt
├── README.md
└── dataset.csv
- Data Cleaning
- Exploratory Data Analysis (EDA)
- Feature Engineering
- Data Preprocessing
- Model Training
- Model Evaluation
- Streamlit Deployment
- Accuracy
- Precision
- Recall
- F1-Score
- Confusion Matrix
git clone https://github.com/yashasvitomar18/Fraud-Detection-Predictor.gitpip install -r requirements.txtstreamlit run fraud_detection.py- Deep Learning-based fraud detection
- Explainable AI (SHAP/LIME)
- REST API integration
- Cloud deployment (AWS/GCP/Azure)
- Real-time transaction monitoring
Yashasvi Tomar