Explainable Artificial Intelligence Model for Electricity Usage Prediction in Nigeria

This study developed a model to predict electricity usage in Nigeria for policy and decision-making. Using XGBoost and LSTM, results showed XGBoost performed better with an RMSE of 0.089, MAPE of 18.51%, and R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> of 0.8, making it the selected model. SHAP-based explainable AI (XAI) provided insights into key factors affecting electricity usage, with the hour of the day being the most significant, followed by the day, month, temperature, wind speed, cloud cover, pressure, and humidity. The model effectively predicts electricity usage and offers valuable insights for improving energy policies and decision-making.