MANGO — Tamil Nadu Electricity Demand Forecasting

An end-to-end Tamil Nadu electricity demand forecasting pipeline built on MANGO, an industrial AI platform with a drag-and-drop flow builder: an 11-block flow around a Random Forest model, with a forecast-vs-actual dashboard.

The Random Forest model scores R² 0.88, MAE 14.19 and MAPE 4.03%. The pipeline is now being extended to gas demand forecasting.

Individual project — built by me during my internship at IIT Madras – Walmart Centre for Tech Excellence (WCTE).

Features

• 11-block flow assembled in MANGO’s drag-and-drop flow builder, ending in a Random Forest forecasting model
• Forecast-vs-actual dashboard comparing predicted and observed demand
• Open datasets published on GitHub under CC BY 4.0

Built with Python, Random Forest, scikit-learn, Streamlit, MANGO.

Source on GitHub