Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

The problem statement is as follows.

Minimize $||g(X\beta)-y||^2$ with respect to $\beta$ where $g(\cdot)$ is some non-linear function, $y$ and $\beta$ are column vectors.

General linear least squares problems are of form $\text{argmin}_\beta\{||X\beta-y||^2\}$ and have plenty of solutions. Simply calling the above problem non-linear least squares produces a myriad of search results and none suits my need. Is there a name for this? Can someone point me in the right direction?

Cheers! = )

share|cite|improve this question
Well, it is a nonlinear least squares problem, in which case there are a number of excellent algorithms (Gauss-Newton, Levenberg-Marquardt, etc.) for finding solutions numerically. See this as well. – J. M. Jun 2 '11 at 4:37
Thank you, the link is very helpful. – Phonon Jun 2 '11 at 14:04
up vote 0 down vote accepted

You might try any numerical analysis book on multidimensional minimization. I don't think the fact that $\beta$ is a vector as opposed to a set of parameters will help. Section 10.4 and on of Numerical Recipes is one source and the obsolete ones are free.

share|cite|improve this answer
Thanks. Looks like a great book. Don't know why I've never heard of it. – Phonon Jun 2 '11 at 14:05

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.