Curve fitting non-linear multi-variable functions

Hello,

I am a bit lost on the concept here, and not sure if this functionality is available in the curve fitting toolbox or machine learning toolbox. I have data from experiments (f(x)) that vary as a function of x. These have been collected by varying a parameter 13 times. I have fitted a non-linear curve (power function) to each of these experiments, and are labelled f(1) to f(13) on the image. Each of these curves share the same independent variable (x), and I am trying to fit the trend of all of these curves into a single function to show how they vary as a function of another parameter (experimental parameter). This creates a multivariable composite function g(f(x),x). Is there any way for me to use an optimization technique in order to fit all of these functions into a single function that show the trend of the curves as I go from 1 to 13? I am quite lost here.

Thank you for any assistance!

Answer 1

[Jyotsna Talluri]

Hi,

You can use the regress function from the statitics and Machine learning toolbox to perform multple linear regression so that obtained plot will be 3D graph with z as a function of x and y

Z=f(x,y);

Refer to the below link

https://www.mathworks.com/help/stats/regress.html