Neural Networks and Logit Models Applied to Commuters’ Mobility in The Metropolitan Area of Milan

Abstract

This chapter presents a study which refers to a supervised model of feedforward neural networks. It briefly describes the main characteristics of neural networks (NNs), especially focusing on the family of two layered feedforward NNs with sigmoidal nonlinearities. The chapter shows the application of a feedforward neural network to the analysis of commuters' mobility in the metropolitan area of Milan (Italy). It also presents case studies concerning the modal split problem, with respect to rail and road transport, of commuters’ flows into Milan from 224 neighbouring municipalities. The analysis, after calibrating the logit model and after training the feedforward NN, investigates the forecasting possibilities of both the approaches. It could be very fruitful to explore further the potentialities and limits of the feedforward neural networks, and of connessionist systems in general above all in relation to the 'conventional' models in order to investigate their promising methodological and computational performance in the area of 'complex' networks.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 11 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.