Combining RBFLN Neural Network and ORESTE Multi-Criteria Technique in Identifying Optimal location for Installation of Financial and Commercial Centers in Urban Spaces (Case Study: Tehran)

Financial and commercial centers (i.e. banks and financial and credit institutes) are considered as an important activity of urban spaces and paying attention to their location and installation site is one of the most important parameters in their success and beneficence. In this study, in order to identify the optimal location for installation of financial and commercial centers the RBFLN neural Network which is a transformed model of Radius Based Function neural Network (RBFNN) was used in combine with ORESTE multi-criteria technique. Two and multi-classes data of economic, commercial, educational, cultural, sanitary, therapeutic, recreational, administrative, population, and transition were entered to the neural network as multi-dimensional vectors based on radius of influence. 69 sample branches and 34 un-optimal points were used for network’s learning. The results indicates the two- classes RBFLN network with 800 repetition times with the least learning and classification error as the most appropriate class in identifying the optimal places for installation of financial and commercial centers (Screening Phase).