In recent years, the Turkish real estate market has emerged as a key driver of economic growth while simultaneously shaping the dynamics of social life. This study employs multivariate analysis methods to classify metropolitan cities and potential metropolitan candidate provinces that exhibit similarities in terms of housing market characteristics, based on 22 socio-economic and sectoral variables influencing the real estate sector. Additionally, the study identifies the metropolitan clusters to which the 10 candidate provinces structurally correspond within this classification framework. To achieve this, conventional classification techniques such as Decision Trees and K-Nearest Neighbors (k-NN) were integrated with artificial intelligence-based methods, including Random Forest (RF) and Support Vector Machines (SVM). The analysis resulted in the categorization of 40 metropolitan and candidate provinces into five distinct groups. Findings indicate that multivariate indicators capturing demographic, economic, and structural differences across metropolitan areas play a critical role in shaping the housing market and guiding strategic urban development decisions. Furthermore, the results highlight that determining metropolitan status solely based on population figures is insufficient and that a more scientific and comprehensive approach—grounded in a broader set of socio-economic and structural indicators yields more meaningful classifications.
Çoban et al. (Tue,) studied this question.