Reproducible Code and Data for Regional Electricity Demand Forecasting in Saudi Arabia

This archive contains the complete reproducible materials supporting the study “Regional Monthly Electricity Load Forecasting in Saudi Arabia: A Comparative Study of SARIMA and Gradient-Boosted Trees.” The repository includes Jupyter notebooks implementing data preprocessing, feature engineering, model estimation, and evaluation for regional electricity load forecasting using monthly data for Saudi Arabia from 2009 to 2019. Forecasting methods include a Seasonal Autoregressive Integrated Moving Average (SARIMA) model and a gradient-boosted decision tree model. Electricity load data are sourced from the KAPSARC Energy Data Portal and are publicly available. The materials are intended to support transparency, replication, and extension of the analysis for academic research and policy applications in energy forecasting and planning.