Séance WA9 - Techniques de programmation mathématique dans les études du comportement en transport / Mathematical Programming Techniques in Transportation Behavior Studies
Jour mercredi, le 06 mai 2009 Salle Banque Scotia Président Fabian Bastin
Présentations
|
10h30- 10h55 |
On the Use of Statistical Tests for Mixed Logit Estimation |
|
Fabian Bastin, Université de Montréal, Informatique et Recherche Opérationnelle, CP 6128, Succ. Centre-Ville, Montréal, Québec, Canada, H3C 3J7 We review classical statistical tests used in discrete choice modelling, and exhibit their limits for advanced models, especially those implying constraints on the parameters. This is for instance the case in some nonparametric formulations. We discuss numerical alternatives, in particular bootstrap techniques applied to the sample population, in order to empirically assess parameters distributions and their significance. |
|
10h55- 11h20 |
Bilevel and Multinomial Logit Models for Hotel Revenue Management |
|
Julio Montecinos, École Polytechnique de Montréal, Mathématiques et génie industriel, C.P. 6079, succursale Centre-ville, Montréal, Quebec, Canada, H3C 3A7 Patrice Marcotte, Université de Montréal, Informatique et recherche opérationnelle, CIRRELT, C.P. 6128, Succ. Centre-ville, Montréal, Québec, Canada, H3C 3J7 Gilles Savard, GERAD, École Polytechnique de Montréal, Mathématiques et génie industriel, C.P. 6079, Succ. Centre-ville, Montréal, Québec, Canada, H3C 3A7 We present a hierarchical optimization model for Hotel Revenue Management where clients’ choices are done minimizing their disutility and following a pseudo-Logit based choice. Hotels' decision variables are prices and “booking limits”. Allowing for Discrete Choice in a context where capacities are restricted is not simple so some redefinitions are done. |
|
11h20- 11h45 |
Dynamic Frameworks for Discrete Choice Models |
|
Cinzia Cirillo, University of Maryland, Civil and Environmental Engineering, 1179 Glenn Martin Hall, College Park, Maryland, USA, 20742 Renting Xu, University of Maryland, Civil and Environmental Engineering, 1179 Glenn Martin Hall, College Park, Maryland, USA, 20742 Discrete choice models are used in transportation planning and modeling, but their theoretical basis have been developed in a static context. We propose to develop an estimation technique for analyzing the impact of technological change on the dynamics of consumer demand. The proposed research presents a model of consumer demand that explicitly accounts for consumers’ expectations of future product quality. |