Comparative ML Model Web App

A modern, interactive web app for credit card fraud detection and model comparison. Built for data science demos, student projects, and practical ML evaluation—no frontend experience required!

🚀 Features

• Data Analysis Tab: Visualize and explore the dataset with clear, beginner-friendly explanations and interactive plots.
• Model Comparison Tab: Compare Random Forest, AdaBoost, XGBoost, and LightGBM side-by-side with metrics, confusion matrices, feature importances, and a comprehensive summary table.
• Make a Prediction Tab: Upload a CSV/Excel row, paste an array, or enter values manually to get predictions from all four models.
• Home Tab: Explains how classic algorithms relate to the project and provides a button to open the original Colab notebook.

🛠️ Tech Stack

• Frontend/Backend: Streamlit
• ML/Analysis: pandas, numpy, scikit-learn, matplotlib, seaborn, plotly, joblib, xgboost, lightgbm

📦 Project Structure

DAA PART B/
├── app/
│   ├── app.py           # Main Streamlit app
│   ├── analysis.py      # Data analysis & visualization logic
│   └── predict.py       # Prediction logic for all models
├── models/              # Saved ML models (.pkl, .json)
├── data/
│   └── creditcard.csv   # Dataset
├── edited_notebook.ipynb# Reference Jupyter notebook
├── requirements.txt     # Python dependencies
└── README.md            # This file

⚡ Quickstart

1. Clone the repo:

   git clone <your-repo-url>
   cd "DAA PART B"

2. Install dependencies:

   pip install -r requirements.txt

3. Run the app:

   streamlit run app/app.py

4. Open in browser:
   • Go to the local URL shown in your terminal (usually http://localhost:8501)

🖼️ Screenshots

Add screenshots of the Data Analysis, Model Comparison, and Prediction tabs here!

📝 Usage

• Data Analysis: Explore the dataset, see summary stats, distributions, correlations, and more.
• Model Comparison: See how each model performs, with clear explanations and a summary table.
• Make a Prediction: Upload a row, paste an array, or enter values to get predictions from all models.
• Home: Learn how classic algorithms power the project and open the Colab notebook with one click.

🤝 Contributing

Pull requests are welcome! For major changes, please open an issue first to discuss what you'd like to change.

🙏 Acknowledgments

• Original dataset
• Streamlit
• Scikit-learn
• XGBoost, LightGBM
• Colab Notebook