GIS-BASED URBAN ROAD NETWORK ACCESSIBILITY MODELING USING MLR, ANN AND ANFIS METHODS

A sustainable transportation system is possible only through an efficient evaluation of transportation network performance. The efficiency of the transport network structure is analyzed in terms of its connectivity, accessibility, network development, and spatial pattern. This study primarily aims to propose a methodology for modeling the accessibility based on the structural parameters of the urban road network. Accessibility depends on the arrangement of the urban road network structure. The influence of the structural parameters on the accessibility is modeled using Multiple Linear Regression (MLR) analysis. The study attempts to introduce two methods of Artificial Intelligence (AI) namely Artificial Neural Networks (ANN) and Adaptive network-based neuro-fuzzy inference system (ANFIS) in modeling the urban road network accessibility. The study also focuses on comparing the results obtained from MLR, ANN and ANFIS modeling techniques in predicting the accessibility. The results of the study present that the structural parameters of the road network have a considerable impact on accessibility. ANFIS method has shown the best performance in modeling the road network accessibility with a MAPE value of 0.287%. The present study adopted Geographical Information Systems (GIS) to quantify, extract and analyze different features of the urban transportation network structure. The combination of GIS, ANN, and ANFIS help in improved decision-making. The results of the study may be used by transportation planning authorities to implement better planning practices in order to improve accessibility.