Accommodating a Flexible Response Heterogeneity Distribution in Choice Models of Human Behavior for Transportation Planning

In this report, the authors propose a general copula approach to accommodate non-normal continuous mixing distributions in multinomial probit (MNP) models. In particular, the authors specify a multivariate mixing distribution that allows different marginal continuous parametric distributions for different coefficients. A new hybrid estimation technique is proposed to estimate the model, which combines the advantageous features of each of the maximum simulated likelihood inference technique and Bhat’s maximum approximate composite marginal likelihood (MACML) inference approach. The effectiveness of this formulation and inference approach is demonstrated through simulation exercises and an empirical application.

Language

  • English

Media Info

  • Media Type: Digital/other
  • Features: Figures; References; Tables;
  • Pagination: 41p

Subject/Index Terms

Filing Info

  • Accession Number: 01666918
  • Record Type: Publication
  • Report/Paper Numbers: D-STOP/2017/131, Report 131
  • Contract Numbers: DTRT13-G-UTC58
  • Files: UTC, TRIS, ATRI, USDOT
  • Created Date: Mar 15 2018 9:37AM