Assessing Users’ Recreational Demand in Urban Parks in Tehran with the Help of the Artificial Neural Network

The advantages of public green spaces are clear. However, it is hard to estimate which parks provide higher standards for users’ (both tourists and citizens) recreational demands. This study provided a model to assess recreational demands in urban parks with the help of the artificial neural network. The aim was to clarify the rules of satisfaction among the users’ recreational demands in urban parks. In 22 districts of the city of Tehran, 104 local urban parks (with a high diversity of the quality of welfare services and design) were selected. Using the user-centered viewpoint, we assessed the recreational demand. A field study from 1395 to 1396 helped to investigate the role of the urban district and park service variables in increasing the demand for urban parks. Results of trained networks showed that the artificial neural network created the best function of topology optimization with a higher coefficient of determination in three categorists of training, validation, and test data. Sensitivity analysis showed that the number of urban district parks, sports areas, cultural areas, and the quality of landscape with a sensitivity coefficient of 183.5, 58.1, 52.7, and 30.4, respectively, had the highest effect on the users’ recreational demand in urban parks. The suggested model would be a decision support system in designing urban parks. Such an approach would help improve urban development based on tourism attractions and would develop urban tourism on a broader scale.