Forecasting and predictive analysis of source-wise power generation along with economic aspects for developed countries

Abstract

This paper presents a comprehensive study on the forecasting and predictive analysis of source-wise power generation, coupled with an examination of economic aspects related to research and development (R&D) in the energy sector for economically significant countries like Australia, the UK, France, the United States, and Germany. This research employs machine learning techniques, including K-Nearest Neighbors (KNN), Decision Tree, Seasonal Autoregressive Integrated Moving Average with Exogenous Factors (SARIMAX), and Autoregressive Integrated Moving Average (ARIMA) models, to provide accurate predictions and insights. Classifier efficiency of the USA, Australia, France, Germany, and the UK is 95.115%, 95.808%, 93.685%, 94.913%, and 93.282%, respectively. Mean Absolute Error (MAE) with KNN and Decision Tree (XGBoost) for the USA, Australia, France, Germany, and UK are 0.578, 0.659, 1.383, 0.738, and 1.02, respectively. This paper also evaluates the relationship between R&D investments in the energy sector and their economic impact, providing policymakers and stakeholders with valuable insights into the long-term benefits of research and sustainable development initiatives.

Keywords

• Renewable energy
• Power generation
• RD&D
• ARIMA
• SARIMAX
• MAE