peter dalgaard gmail.com> writes:
>
> You are being over-optimistic with your starting values, and/or
> with constrains on the parameter space.
> Your fit is diverging in sigma for some reason known
> only to nonlinear-optimizer gurus...
>
> For me, it works either to put in an explicit
> c
You are being over-optimistic with your starting values, and/or with constrains
on the parameter space.
Your fit is diverging in sigma for some reason known only to
nonlinear-optimizer gurus...
For me, it works either to put in an explicit constraint or to reparametrize
with log(sigma).
E.g.
Hi everyone,
I have a problem with maximum-likelihood-estimation in the following
situation:
Assume a functional relation y = f(x) (the specific form of f should be
irrelevant). For my observations I assume (for simplicity) white noise,
such that hat(y_i) = f(x_i) + epsilon_i, with the epsilon_i
3 matches
Mail list logo
