MODELLING OF DAILY ACTIVITY SCHEDULE OF WORKERS USING UNSUPERVISED MACHINE LEARNING TECHNIQUE

Abstract

Travel demand models are used to replicate the real world travel demand and to predict the future travel demand. A behavioural oriented approach in travel demand analysis is provided by activity based travel demand modelling and it provides a better understanding of the travel behaviour of an individual. A sequential modelling approach using econometric models is commonly used in activity based travel modelling. In this method, error obtained in one model will be carried forward to the second model and so on. Hence chances of accumulation of error are more at the final stage, when the models are used for prediction. Hence an attempt is made in this study to replace the sequential econometric modelling approach with simultaneous modelling approach using an unsupervised machine learning technique. Three stage Neural Network modelling used in this study replaces sixteen stage econometric models. The predictive accuracy of all the output parameters was compared in both the modelling approaches. Results shows that Artificial Neural Network (ANN) results outperform econometric models. The decrease in percentage error ranges from 2.22% to 27.17%.