Modeling and Prediction of the Aesthetics of Urban Parks Based on Landscape Complexity Criterion

People are looking for new spaces that they can discover, and keep distance from dull and repetitive spaces, to create a place that is compatible with today's complexity and specific desires. The purpose of this study is to model the complexity of landscape in urban parks using artificial neural network in order to predict the complexity of landscape in urban parks and determination of the effect of different landscape variables on it. The present study was carried out in ten parks with an area of more than 10 hectares in Tehran city (Saei, Mellat, Nahj al-Balagheh, Iran Zaman, Laleh, Abo Atash, Taleghani, Jamshidieh, Gheitarieh and Niavaran) and in four municipality districts (1, 2, 3, and 6). The purpose of this study was to evaluate the complexity of landscape in urban park using a combination of user-based approach and artificial neural network modeling using 17 objective elements of landscape. According to the results, the model with 17-14-1 structure (17 input variables, 14 neurons in hidden layer and one output variable) according to the highest value of coefficient of determination in the three categories of training, validation and test datasets equal with 0.93, 0.85, and 0.87 created the best optimization performance. Accordingly, hard surfaces ratio, the mean of trees diameter and buildings, with the coefficients of 0.21, 0.15, and 0.12, respectively, are the most influential factors on the complexity of landscape in Tehran city parks, respectively. The presented model in this study is known as a decision support system in engineering design of urban parks and enables prediction of the complexity of landscape according to the environmental variables. .