Re: parameter estimation, special form

slitvinov_at_gmail.com
Date: 03/16/05 

Date: 15 Mar 2005 16:01:54 -0800

Peter Spellucci wrote:

> for nonlinear least squares codes see
> http://plato.la.asu.edu/topics/problems/nlolsq.html
> hth
> peter

That's a good point! I found

'netlib/opt/varpro separable problem, where Phi depends linearly on
a part of the parameters (e.g. sum_i a_i*exp(b_i*t)). These can be
treated in a special way, leaving a nonlinear problem of reduced
dimension'

I have one of such parameters (p_3).

Relevant Pages