This paper analyses the purchase behaviour for conventional and alternative fuel cars, using Italian stated preference discrete choice data. We propose a flexible Hierarchical Bayesian Mixed Logit (HBML) model that permit us to take into account of possible dependences of the car attribute random parameters on individual characteristics, like age and gender. Moreover, alternative-specific parameters and correlation across alternatives have been easily added to the model. We carried out a survey of the literature on vehicle purchase choice selecting applications of discrete choice models in which a Bayesian approach was adopted. It reveals that our study seems to be the first application of HBML models to analyse this type of stated choices. Moreover, in order to approximate the joint posterior distribution of both the model parameters and hyper-parameters, in this paper we use the most efficient Hamiltonian Monte Carlo sampler, instead of considering the more traditionally Markov Chain Monte Carlo (MCMC) methods as e.g. Gibbs sampler. The modelling results show the usefulness of the proposed method.

Hierarchical Bayes Mixed logit modelling for purchase car behaviour

CARMECI, GAETANO;
2015-01-01

Abstract

This paper analyses the purchase behaviour for conventional and alternative fuel cars, using Italian stated preference discrete choice data. We propose a flexible Hierarchical Bayesian Mixed Logit (HBML) model that permit us to take into account of possible dependences of the car attribute random parameters on individual characteristics, like age and gender. Moreover, alternative-specific parameters and correlation across alternatives have been easily added to the model. We carried out a survey of the literature on vehicle purchase choice selecting applications of discrete choice models in which a Bayesian approach was adopted. It reveals that our study seems to be the first application of HBML models to analyse this type of stated choices. Moreover, in order to approximate the joint posterior distribution of both the model parameters and hyper-parameters, in this paper we use the most efficient Hamiltonian Monte Carlo sampler, instead of considering the more traditionally Markov Chain Monte Carlo (MCMC) methods as e.g. Gibbs sampler. The modelling results show the usefulness of the proposed method.
2015
9788883037184
File in questo prodotto:
File Dimensione Formato  
Carmeci_Valeri_Auto_elettrica.pdf

accesso aperto

Descrizione: pdf editoriale
Tipologia: Documento in Versione Editoriale
Licenza: Digital Rights Management non definito
Dimensione 1.06 MB
Formato Adobe PDF
1.06 MB Adobe PDF Visualizza/Apri
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11368/2870713
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact