Term Deposit Subscription Predictor: SMOTE + Random Forest + Streamlit
A full-stack machine learning project that predicts whether customers will subscribe to a term deposit based on banking campaign data. Built with Python, Random Forests, SMOTE, and deployed in an interactive Streamlit app.
📊 Project Summary
- Cleaned and encoded banking campaign data with over 40k records.
- Trained a Random Forest classifier and improved results with SMOTE balancing.
- Visualized top predictors, evaluation metrics, and customer segments.
- Deployed an interactive Streamlit app with navigation and batch prediction.
🧰 Tools Used
- Pandas, NumPy, Matplotlib, Seaborn
- Scikit-learn (train/test split, classification, SMOTE)
- Streamlit for dashboard deployment
📷 Dashboard Screenshots
📓 Jupyter Notebook
Download the notebook used for modeling, EDA, and Streamlit setup.
Download Notebook (.ipynb)