Railway Passenger Train Delay Prediction via Neural Network Model

This article presents an artificial neural network (ANN) model with high accuracy that can be used to predict the delay of passenger trains in Iranian Railways. The authors use three different methods to define inputs: normalized real number, binary coding, and binary set encoding. The authors report on their investigations of three different strategies (quick method, dynamic method, and multiple method) for designing the neural network. They also divide the existing passenger train delays data set into three subsets called training set, validation set, and testing set. The study compared the results of three different data input methods and three different architectures with each other and with some common prediction methods such as decision tree and multinomial logistic regression. As part of the comparison of different neural networks, the authors focus on training time and accuracy of neural networks on test data set and network size. Finally, to make a fair comparison among all models, the authors sketched a time-accuracy graph. They conclude that their proposed model has higher accuracy than earlier methods. This tool is necessary because being able to predict delays enables railway operators to maintain a suitable timetable and minimize overall delays, errors, and problems in planning.

Language

  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01492198
  • Record Type: Publication
  • Files: TRIS
  • Created Date: May 8 2013 3:21PM