Enhancing discrete choice models with representation learning

被引:65
|
作者
Sifringer, Brian [1 ]
Lurkin, Virginie [2 ]
Alahi, Alexandre [1 ]
机构
[1] Ecole Polytech Fed Lausanne EPFL, Visual Intelligence Transportat Lab Vita, Lausanne, Switzerland
[2] Eindhoven Univ Technol, Dept Ind Engn & Innovat Sci, NL-5600 MB Eindhoven, Netherlands
关键词
Discrete choice models; Neural networks; Utility specification; Machine learning; Deep learning; ARTIFICIAL NEURAL-NETWORKS; MULTINOMIAL LOGIT-MODELS; TRAVEL MODE; UTILITY FUNCTION; MACHINE; SPECIFICATION; BEHAVIOR; LEGIT;
D O I
10.1016/j.trb.2020.08.006
中图分类号
F [经济];
学科分类号
02 ;
摘要
In discrete choice modeling (DCM), model misspecifications may lead to limited predictability and biased parameter estimates. In this paper, we propose a new approach for estimating choice models in which we divide the systematic part of the utility specification into (i) a knowledge-driven part, and (ii) a data-driven one, which learns a new representation from available explanatory variables. Our formulation increases the predictive power of standard DCM without sacrificing their interpretability. We show the effectiveness of our formulation by augmenting the utility specification of the Multinomial Logit (MNL) and the Nested Logit (NL) models with a new non-linear representation arising from a Neural Network (NN), leading to new choice models referred to as the Learning Multinomial Logit (L-MNL) and Learning Nested Logit (L-NL) models. Using multiple publicly available datasets based on revealed and stated preferences, we show that our models outperform the traditional ones, both in terms of predictive performance and accuracy in parameter estimation. All source code of the models are shared to promote open science. (C) 2020 Elsevier Ltd. All rights reserved.
引用
收藏
页码:236 / 261
页数:26
相关论文
共 50 条
  • [41] ON DISCRETE CHOICE MODELS OF HOUSING DEMAND
    BORSCHSUPAN, A
    PITKIN, J
    JOURNAL OF URBAN ECONOMICS, 1988, 24 (02) : 153 - 172
  • [42] Risk, uncertainty and discrete choice models
    Andre de Palma
    Moshe Ben-Akiva
    David Brownstone
    Charles Holt
    Thierry Magnac
    Daniel McFadden
    Peter Moffatt
    Nathalie Picard
    Kenneth Train
    Peter Wakker
    Joan Walker
    Marketing Letters, 2008, 19 : 269 - 285
  • [43] Generalized Reverse Discrete Choice Models
    Sanjog Misra
    Quantitative Marketing and Economics, 2005, 3 : 175 - 200
  • [44] Structure and uncertainty in discrete choice models
    Signorino, CS
    POLITICAL ANALYSIS, 2003, 11 (04) : 316 - 344
  • [45] A FLEXIBLE CLASS OF DISCRETE CHOICE MODELS
    DALAL, SR
    KLEIN, RW
    MARKETING SCIENCE, 1988, 7 (03) : 232 - 251
  • [46] Risk, uncertainty and discrete choice models
    de Palma, Andre
    Ben-Akiva, Moshe
    Brownstone, David
    Holt, Charles
    Magnac, Thierry
    McFadden, Daniel
    Moffatt, Peter
    Picard, Nathalie
    Train, Kenneth
    Wakker, Peter
    Walker, Joan
    MARKETING LETTERS, 2008, 19 (3-4) : 269 - 285
  • [47] Nonparametric continuous/discrete choice models
    Newey, Whitney K.
    INTERNATIONAL ECONOMIC REVIEW, 2007, 48 (04) : 1429 - 1439
  • [48] ON COMBINING DISCRETE CHOICE AND ASSIGNMENT MODELS
    Larsen, Odd I.
    Jansson, Kjell
    Lang, Harald
    TRANSPORTATION AND URBAN SUSTAINABILITY, 2010, : 299 - 306
  • [49] On the variability of hybrid discrete choice models
    Bahamonde-Birke, Francisco J.
    de Dios Ortuzar, Juan
    TRANSPORTMETRICA A-TRANSPORT SCIENCE, 2014, 10 (01) : 74 - 88
  • [50] Assisted specification of discrete choice models
    Ortelli, Nicola
    Hillel, Tim
    Pereira, Francisco C.
    de Lapparent, Matthieu
    Bierlaire, Michel
    JOURNAL OF CHOICE MODELLING, 2021, 39