Weibit model

14 views
Skip to first unread message

RONNY MARCELO ALIAGA MEDRANO

unread,
Jul 3, 2024, 5:49:46 PMJul 3
to Biogeme
Dear Professor Bierlare,
Recently, i read about the Weibit discrete choice model. Could be possible to program it with biogeme, following the process where the weibit model is estimated using the logit estimation module based on a log transformation of the utility function? as propossed in Fosgerau and Bierlaire, 2009, Section 4.1
As for the modeling application, I believe the weibit model can be applied to latent class and hybrid choice models, given that a suitable error distribution is assumed for the latent variables, such that all the random variables are defined in the positive domain and the multiplicative error structure of weibit model can be retained. Is it correct?
Best regards,

Michel Bierlaire

unread,
Jul 4, 2024, 9:34:07 AMJul 4
to ronnym...@ufg.br, Michel Bierlaire, Biogeme


> On 3 Jul 2024, at 20:55, RONNY MARCELO ALIAGA MEDRANO <ronnym...@ufg.br> wrote:
>
> Dear Professor Bierlare,
> Recently, i read about the Weibit discrete choice model. Could be possible to program it with biogeme, following the process where the weibit model is estimated using the logit estimation module based on a log transformation of the utility function?

This is a possibility. It allows to exploit the logit implementation, which is optimized for speed.
But it is not necessary. In Biogeme, you simply need to write the formula for the contribution of the log likelihood function of one observation. It does not have to be logit.

> as propossed in Fosgerau and Bierlaire, 2009, Section 4.1
> As for the modeling application, I believe the weibit model can be applied to latent class and hybrid choice models, given that a suitable error distribution is assumed for the latent variables, such that all the random variables are defined in the positive domain and the multiplicative error structure of weibit model can be retained. Is it correct?

Yes.

> Best regards,
>
> --
> You received this message because you are subscribed to the Google Groups "Biogeme" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to biogeme+u...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/biogeme/aa8ab085-d4c6-4e57-bd97-341a5bd0e9edn%40googlegroups.com.

Michel Bierlaire
Transport and Mobility Laboratory
School of Architecture, Civil and Environmental Engineering
EPFL - Ecole Polytechnique Fédérale de Lausanne
http://transp-or.epfl.ch
http://people.epfl.ch/michel.bierlaire

Reply all
Reply to author
Forward
0 new messages