ANALYSIS OF CITIES DATA USING PRINCIPAL COMPONENT INPUTS IN AN ARTIFICIAL NEURAL NETWORK

Abstract

Trip rate is one of the transportation planning parameter. It is difficult to relate the amount of trips that originate in a study area and the amount of trips attracted towards that study area by conducting surveys regularly. The cost and time for each survey is not affordable. So considering Travel parameters and Land-use Parameters of an area relationship is established using Artificial Neural Network (ANN) against Trip Rates of that area. Involving more number of parameters has made computations to increase the complexity in analysis. So the data has been reduced in dimension using Principal Component Analysis and then Processed in an Artificial Neural Network. The original input data along with principal components (6PC, 5PC, 4PC and 3PC) data as input to the Artificial Neural Network (ANN) has been processed separately. The analysis has shown that 6PC as input to Artificial Neural Network(ANN) is yielding better explanation between independent and dependent variables (Trip Rate(all modes) and Trip Rate(motorised)).