choice-metrics.com

Dear all, I am trying to understand how to estimate an efficient design for a mixed logit model.

I reviewed NGENE appendix section and the paper Bliemer, M. C., & Rose, J. M. (2010). Construction of experimental designs for mixed logit models allowing for correlation across choice observations. Transportation Research Part B: Methodological, 44(6), 720-734.

In both, the Hessian formula is shown in general terms. I have been struggling to derive the Hessian myself in more detail.

Can you recommend me a text where a detailed mathematical derivation for the Hessian is presented?

Thanks

Cristobal

Unfortunately not really - this took us years to derive, program, test, etc. I'm not aware that it has been reported elsewhere.

John