Regression in matlab to estimate parameters

Hi,

I want to use regression to fit parameters to a function I define. I tried using fminsearch in matlab, and it doesn't converge. However, in excel solver the GRG Nonlinear method works very well. Is there any similar function in matlab for this purpose? (With experimental data, I used the function to predict the error. I used fminsearch to minimize sum(deviation^2). )

Answer 1

[the cyclist]

If you have the Statistics and Machine Learning Toolbox, you can use the nlinfit function to fit an arbitrary function.

The documentation has some examples. If you search by username and "nlinfit", you will also find a couple I have posted here.

[Bharat]

Thank you. Just checked and it works!

[the cyclist]

The best form of thanks is to upvote and/or accept an answer, which rewards the contributor and can point future users to useful answers.

[Brendan Hamm]

I'll also point you to the newer and better version of nlinfit, called fitnlm. There is way more functionality surrounding the Statistics & Machine Learning Toolbox's objects.