A decomposition method for estimating recursive logit based route choice models

Fosgerau et al (2013) recently proposed the recursive logit (RL) model for route choice problems that can be consistently estimated and easily used for prediction without any sampling of choice sets. Its estimation however requires solving many large-scale systems of linear equations, which can be computationally costly for real data sets. We design a decomposition (DeC) method in order to reduce the number of linear systems to be solved, opening the possibility to estimate more complex RL based models, such as, mixed RL models. We illustrate the approach on two mixed RL specifications, one using random coefficients and one incorporating error components associated with subnetwork (Frejinger and Bierlaire, 2007). The models are estimated on a real network with more than 3000 nodes and 7000 links, and a cross-validation study is performed. The results suggest that the DeC method significantly speeds up the estimation of the RL model and allows to estimate the mixed RL models in a reasonable time. The mixed RL model yields sensible parameter estimates and the fit and prediction are significantly better than the RL model.


  • English

Media Info

  • Pagination: 23p

Subject/Index Terms

Filing Info

  • Accession Number: 01593974
  • Record Type: Publication
  • Source Agency: ARRB
  • Report/Paper Numbers: CIRRELT-2015-66
  • Files: ATRI
  • Created Date: Mar 21 2016 11:47AM