A review of geomarketing methods for optimal retail location selection

The article presents a comprehensive analysis of modern geomarketing approaches to solving the problem of optimal location selection for retail enterprises. A systematic review and comparative analysis of the evolution of models and methods are conducted: from classical location theories (central place theory, law of retail gravitation, principle of minimum differentiation) to advanced machine learning-based algorithms. The paper details traditional models (Huff model), the role of geographic information systems (GIS), multi-criteria decision analysis methods and their limitations. Special attention is paid to the application of regression models, ensemble methods (random forest, gradient boosting), and neural networks in geomarketing. The key types of geospatial data, their sources, and model performance evaluation metrics are analyzed. The article systematizes the accumulated knowledge and demonstrates a steady trend towards the use of complex, data-driven methods for strategic decision-making in retail.