Bayesian estimation of hybrid choice models

Hi all,

I am trying to model a hybrid choice model (HCM) using Bayesian estimation.

I am a beginner in PyStan and I did not find any example script which estimates any choice model.

Please guide or direct me to the specified documentation or similar example so that I can use that to model HCM.


Ishant Sharma

Hey @Ishant_Sharma welcome to the community! If you write out a model (in words or maths) for your scenario then we can help you build a Stan coding of it here. Would be delighted to help.

1 Like

Hi @emiruz, Thanks.

I have added the mathematical form for model relationships and maximum simulated likelihood (MSL) estimation in the attached PDF. I want to replace MSL with Bayesian estimation.

Please let me know if this will work.

HCM_simultaneous.pdf (105.9 KB)

Cool thanks. It’s quite a generic model so it’s quite laborious to write up without further details. Do you think you could maybe provide an example under a specific design? Ideally, you’d provide a fake data generating function that the model can aim to discover the parameters of. That’d give us a good basis to help from.

Otherwise some of the other members may have specific experience with choice models to volunteer!