TRIP DISTRIBUTION FORECASTING WITH MULTILAYER PERCEPTRON NEURAL NETWORKS: A CRITICAL EVALUATION

This study compares the performance of multilayer perceptron neural networks and maximum-likelihood doubly-constrained models for commuter trip distribution. Our experiments produce overwhelming evidence at variance with the existing literature that the predictive accuracy of neural network spatial interaction models is inferior to that of maximum-likelihood doubly-constrained models with an exponential function of distance decay. The study points to several likely causes of neural network underperformance, including model non-transferability, insufficient ability to generalize, and reliance on sigmoid activation functions, and their inductive nature. It is concluded that current perceptron neural networks do not provide an appropriate modeling approach to forecasting trip distribution over a planning horizon for which distribution predictors (number of workers, number of residents, commuting distance) are beyond their base-year domain of definition.

[1]  Morton E. O'Kelly,et al.  Spatial Interaction Models:Formulations and Applications , 1988 .

[2]  Murray Smith,et al.  Neural Networks for Statistical Modeling , 1993 .

[3]  Carl Amrhein,et al.  The effect of data aggregation on a poisson model of Canadian migration , 1989 .

[4]  Armando Freitas da Rocha,et al.  Neural Nets , 1992, Lecture Notes in Computer Science.

[5]  Michael Batty,et al.  The Calibration of Gravity, Entropy, and Related Models of Spatial Interaction , 1972 .

[6]  Paul B. Slater.  Hierarchical Internal Migration Regions of France , 1976, IEEE Transactions on Systems, Man, and Cybernetics.

[7]  Brian D. Ripley,et al.  Pattern Recognition and Neural Networks , 1996 .

[8]  Michel Verleysen,et al.  An optimized RBF network for approximation of functions , 1994, ESANN.

[9]  J. Nazuno.  Haykin, Simon. Neural networks: A comprehensive foundation, Prentice Hall, Inc. Segunda Edición, 1999 , 2000 .

[10]  Simon Haykin,et al.  Neural Networks: A Comprehensive Foundation , 1998 .

[11]  R. W. Thomas,et al.  Modeling in Geography: A Mathematical Approach , 1980 .

[12]  Shinya Kikuchi,et al.  A METHOD TO ESTIMATE TRIP O-D PATTERNS USING A NEURAL NETWORK APPROACH. , 1993 .

[13]  Thomas H. Wonnacott,et al.  Introductory Statistics , 2007, Technometrics.

[14]  Peter Nijkamp,et al.  Neural Networks in Transport Applications , 1998 .

[15]  Andrew W. Evans,et al.  The calibration of trip distribution models with exponential or similar cost functions , 1971 .

[16]  A. Stewart Fotheringham,et al.  Goodness-of-fit statistics , 1987 .

[17]  S. Hyakin,et al.  Neural Networks: A Comprehensive Foundation , 1994 .

[18]  Lourdes Calderón.  Neural networks as pattern recognition systems , 1996 .

[19]  Geoffrey E. Hinton,et al.  Learning representations by back-propagating errors , 1986, Nature.

[20]  Michael Bacharach,et al.  Biproportional matrices & input-output change , 1970 .

[21]  Mark Dougherty,et al.  A REVIEW OF NEURAL NETWORKS APPLIED TO TRANSPORT , 1995 .

[22]  M Batty,et al.  Spatial Aggregation in Gravity Models. 1. An Information-Theoretic Framework , 1982, Environment & planning A.

[23]  Michael Green.  Biproportional Matrices and Input‐Output Change , 1971 .

[24]  Raúl Rojas,et al.  Neural Networks - A Systematic Introduction , 1996 .

[25]  Carl Amrhein,et al.  The Effect of Data Aggregation on a Poisson Regression Model of Canadian Migration , 1992 .

[26]  Christopher M. Bishop,et al.  Neural networks for pattern recognition , 1995 .

[27]  Ardeshir Faghri,et al.  APPLICATIONS OF ARTIFICIAL NEURAL NETWORKS TO INTELLIGENT VEHICLE-HIGHWAY SYSTEMS , 1994 .

[28]  William R. Black,et al.  Spatial interaction modeling using artificial neural networks , 1995 .

[29]  Manfred M. Fischer,et al.  ARTIFICIAL NEURAL NETWORKS: A NEW APPROACH TO MODELING INTERREGIONAL TELECOMMUNICATION FLOWS* , 1994 .

[30]  Stan Openshaw.  Modelling spatial interaction using a neural net , 1993 .