A Method for Assessing and Predicting Urban Expansion

Abstract. This study introduces an integrated framework for assessing and predicting urban expansion, illustrated through the case of Urmia, utilizing artificial intelligence (AI), multi-criteria decision-making (MCDM), and GIS. The multi-layer perceptron (MLP) analysis predicted the total built-up area to be 134.29 km², of which 82.32 km² is currently developed, leaving 51.97 km² available for future expansion. The combined application of Support Vector Machine (SVM), Analytic Network Process (ANP), and MLP techniques demonstrated high accuracy in land use mapping and growth predictions, highlighting the effectiveness of machine learning in addressing urban complexities. Key factors such as geology and proximity to roads were identified as significant drivers of urban growth, indicating the need for localized strategies in future research. This methodology can be applied to similar urban contexts, offering a data-driven approach to sustainable urban planning and effective management of rapid urbanization.