<![CDATA[Effective energy system planning requires energy demand projections that are reliable, stable, and easy to implement, particularly under conditions of limited historical data and computational resources. However, many existing artificial intelligence–based forecasting approaches are highly complex, difficult to interpret, and time-consuming to develop, which reduces their practicality for students and researchers who aim to focus on solution-oriented energy system analysis. This paper proposes a simple yet reliable energy demand projection framework by combining statistical time series modeling and machine learning methods, namely Auto Regressive Integrated Moving Average (ARIMA) and Ridge Regression. The ARIMA model is employed to capture the temporal dynamics of energy consumption and to construct a business-as-usual (BAU) scenario based on historical trends. The ARIMA projections are subsequently used as inputs for the Ridge Regression model, which captures the multivariate relationships between energy demand and correlated socio-economic factors. The results indicate that ARIMA effectively represents historical consumption patterns but tends to produce conservative projections. In contrast, Ridge Regression provides more stable and robust estimates under conditions of high multicollinearity and limited sample size. The integration of these two methods results in an efficient, interpretable, and easily reproducible modeling framework. The proposed approach is intended to help students and researchers reduce the time required for energy demand forecasting, allowing them to focus more on solution development and sustainable energy system planning.]]>
