Re: Constrained non linear regression using ML

[Jaroslav Hajek]

On Tue, Mar 23, 2010 at 9:33 AM, Corrado <address@hidden> wrote:
> Sorry, I am really lost here ....
>
> Jaroslav Hajek wrote:
>>>
>>> Yes and no. It's not a simple random variable, it's dependent on an
>>> n-dimensional vector p: Y = Y(p). In statistics this is usually called
>>> a "random field".
>>>
>
> What is p, and what is n?

p is from your original parametric model:
Y(p) = 1-exp(-(k0+k'*p)) + Yerr

where Yerr is a random variable (the error component), i.e. Y is a sum
of a function of p and an independent random variable.


-- 
RNDr. Jaroslav Hajek, PhD
computing expert & GNU Octave developer
Aeronautical Research and Test Institute (VZLU)
Prague, Czech Republic
url: www.highegg.matfyz.cz