Hey members,
I would like to generate a 2-point bayesian D-optimal design for the probit model as follows:
- The model is the standard probit model i.e. Y=Φ((x-μ)/σ)
- My priors are: μ~Ν(1000, 50^2), σ=0.05*μ*α, log(α)~Ν(0,log(2)).
How should I go about it?
Thanks!
Liran
