Biogeme in a nutshell

Biogeme (Bierlaire's Optimization Toolbox for GEV Model Estimation) is an object-oriented software package designed for the maximum likelihood estimation of Generalized Extreme Value (GEV) models.

The current version of Biogeme has the following main features:

• GEV models: the following models can be estimated:
  • Multinomial Logit models,
  • Nested logit models,
  • Cross-Nested logit models,
  • Network-GEV models.
• Binary probit models can be estimated.
• Continuous and discrete mixtures of GEV models: normally distributed and uniformly distributed random parameters can appear in the utility functions, as well as random parameters with discrete distributions. In the presence of parameters with continuous distributions, Simulated Maximum Likelihood is performed, using either pseudo-random numbers, quasi-random numbers generated from Halton sequences, or the Modified Latin Hypercube Sampling strategy.
• Panel data: it is possible to estimate individual-specific random parameters.
• No scaling issue: An important feature of Biogeme is that the specification of the utility function is completely independent from the distribution of the error term. It has the advantage that the user does not need to worry about scaling problems when constraining parameters appearing in different nests of a nested or cross-nested logit model, for instance. So the issues raised by Koppelman and Wen (1998a) and Koppelman and Wen (1998b) are not relevant in this context.
• Nonlinear utility functions: the utility functions do not need to be linear-in-parameters, like in most standard packages. A wide range of nonlinear specifications are allowed. No derivative is requested from the user. As a byproduct, log-normally distributed parameters can also be specified.

Conditions of use

BIOGEME is distributed free of charge. We ask each user

• to register to Biogeme's users group, and
• to mention explicitly the use of the package when publishing results, using the following reference:

  Bierlaire, M. (2003). BIOGEME: A free package for the estimation of discrete choice models , Proceedings of the 3rd Swiss Transportation Research Conference, Ascona, Switzerland.

  We suggest also to cite the tutorial and the URL biogeme.epfl.ch

  Bierlaire, M. (2008). An introduction to BIOGEME Version 1.7, biogeme.epfl.ch

Disclaimer

This software is provided free of charge and WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. Nor myself neither my employer (EPFL) shall be liable for any damage of any kind in connection with the use of this software.