Boston Housing Price Prediction

A machine learning project to train and serve a Boston Housing regression model using Streamlit.

What is implemented

• Reproducible training pipeline (src/train.py) that:
  • reads data/data.csv
  • trains a RandomForestRegressor
  • evaluates on a holdout test set (MAE / RMSE / R²)
  • saves artifacts to models/
• Prediction helper (src/predict.py) for consistent feature ordering and model loading.
• Streamlit UI (app.py) with:
  • form controls for all model features
  • one-click prediction
  • display of last recorded training metrics

Project structure

.
├── app.py
├── data/
│   ├── data.csv
│   └── details.png
├── models/                # generated after training
├── src/
│   ├── predict.py
│   └── train.py
├── sample.ipynb
└── requirements.txt

Quick start

1. Create and activate a virtual environment (recommended).
2. Install dependencies:

pip install -r requirements.txt

3. Train the model:

python src/train.py

4. Run the app:

streamlit run app.py

Notes

• This dataset is commonly used for education/demo purposes; it also has known ethical limitations in modern ML fairness contexts.
• For production use, add stronger validation, model versioning, CI checks, and monitoring.