The Relationship between Area of Arterial Roads and Related Factors in City Plan by using Artificial Neural Network (ANN)

Abstract

Road network plan is an important part in City plan. This research aims to develop guidelines in order to make road area in road network are agreeable with land use plan. In order to apply Artificial neural network in this study, 18 city plans were collected as samples. There are 10 variables in each city plan. Major road area in road network was selected as dependent variable and the less were selected as independent variables. There are only 2 independent variables were selected from Variable selection analysis as number of population and residential area. By using Split test, the accuracy analysis shows acceptable sample at only 25% and 67% for the first and second round respectively. Anyway, the overall accuracy analysis shows acceptable samples only 43%. Finally, Artificial neural network (ANN) may not appropriate with determining major road network area in city plan.