Discover MakerZone

MATLAB and Simulink resources for Arduino, LEGO, and Raspberry Pi

Learn more

Discover what MATLAB® can do for your career.

Opportunities for recent engineering grads.

Apply Today

How to estimate two non-linear regression models jointly in MATLAB

Asked by R on 21 Aug 2014
Latest activity Commented on by Star Strider on 22 Aug 2014

Hi, I am new to this forum and have little experience with MATLAB. I would appreciate your help with the following question.

Is there a function or a toolbox in MATLAB that allows to estimate non-linear regression model parameters jointly?

I have two non-linear regression models to estimate jointly.

Thanks, R

0 Comments

R

Tags

Products

No products are associated with this question.

1 Answer

Answer by Star Strider on 21 Aug 2014

I do not understand what you mean by ‘jointly’, but there are three possible functions you can use: nlinfit (Statistics Toolbox), lsqcurvefit (Optimization Toolbox) or (with an extra line of code) the core MATLAB function fminsearch.

4 Comments

Star Strider on 22 Aug 2014

I still have no idea what you want to do.

Again:

  1. What are your independent (‘x’) and dependent (‘y’) variables with respect to the equations you posted?
  2. Are you fitting one dependent variable or two?
Star Strider on 22 Aug 2014

I can help you with the nonlinear regression functions that I mentioned in my original Answer, but you need to code your objective functions in terms of the parameters and independent variables, as described in the documentation for nlinfit or lsqcurvefit. Only you know what your data are, and how you want to code your functions. You have to make the initial effort.

Star Strider on 22 Aug 2014

The lsqcurvefit function can fit matrix dependent variables, but since it is impossible for me to figure out what your independent and dependent variables are, I can’t help you to code them.

It looks as though they are nonlinear equations log-transformed to make them linear. (This is definitely not a recommended practice.)

Post the original equations (not the log-transformed versions), specify your independent and dependent variables, make a sincere attempt at coding them, and I can help you fit them with lsqcurvefit.

Star Strider

Contact us