The application is publicly available on Streamlit Community Cloud at the following link:

🐍 📄 A model to predict individual income thresholds in the US (<=50K or >50K) using synthetically generated data via CTGAN and classified with XGBoost.

To simulate a realistic dataset of US adults and build a machine learning model capable of predicting annual income, then visualize everything interactively using Streamlit.

• 📊 Original Dataset: UCI Adult Dataset
• 🧬 Synthetic Data Generator: CTGAN
• 🤖 Classifier: XGBoost
• 🔍 Model Explainability: SHAP values
• 🌐 Interactive Dashboard: Streamlit

The Streamlit app includes the following sections:

1. Synthetic dataset generated with CTGAN
2. XGBoost model performance report
3. SHAP plots to explain feature importance
4. Download button to export synthetic data

👉 Try it now by clicking the badge above!

Streamlit_USAdult_Income/ ├── data/ # Original and synthetic data ├── outputs/ # SHAP plots and model results ├── streamlit_app/ # Streamlit app code │ └── app.py ├── requirements.txt # Python dependencies └── README.md

git clone https://github.com/DanteTrb/Streamlit_USAdult_Income.git
cd Streamlit_USAdult_Income
pip install -r requirements.txt
streamlit run streamlit_app/app.py


👤 Author
Dante Trabassi
LinkedIn: https://www.linkedin.com/in/dante-trabassi-663b3718b/