Applications of Machine Learning for Energy and Buildings in MENA Area: A Review Paper

Artificial intelligence (AI) overall and machine learning (ML) in particular have been increasingly popular in recent years. Two reviews are included in the paper: (1) A quick review of the four main machine learning (ML) approaches, which include artificial neural networks (ANN), support vector machines (SVM), Gaussian-based regressions (i.e., Gaussian process regression (GPR) or Gaussian mixture models (GMM)), and clustering (such as k-means and k-shape clustering algorithms) that have been widely used in forecasting and enhancing building energy performance. (2) A review of possible applications for each technique in the Middle East and North Africa (MENA) region. Based on the two reviews, this paper proposes two layouts to assist everyone (especially MENA area engineers with limited experience in the building energy and machine learning fields) in selecting the most appropriate machine learning approach for a given application and explaining the prerequisites for each one to improve performance.