I want to curve fit an output with multiple inputs?

I have a set of data that has multiple inputs with one output. I'd like to know how to get an equation that estimates my output using Matlab. Is there a tutorial or someone i can speak with to know if the product does this before purchasing.

[Adam Danz]

There are lots of tools built into Matlab that are designed to solve curve fitting and parameter estimation problems as you're describing. This page will introduce some of those tools to you.

https://www.mathworks.com/products/curvefitting.html

Here's an example function that fits nonlinear curves.

https://www.mathworks.com/help/optim/ug/lsqcurvefit.html

[dpb]

Attach one dataset and an idea of the functional you want to fit and somebody will probably show an actual case...

[Walter Roberson]

If you have more than 2 independent variables, then the curve fitting toolbox is not set up to handle that.

Note that the curve fitting toolbox is for the case where you have a model with parameters and you want to find the parameters. If you have several candidate models, you can evaluate each of them in turn and find the one that gives the best fit. However, if you do not have a finite list of models to try, then curve fitting can never tell you what the equation "really" is.

If you have a model but with more than two independent variables, then lsqnonlin can be used, amongst other techniques

There are at least two other major approaches:

• using neural networks to approximate a function, using the Deep Learning Toolbox
• Systems Identification, using the Systems Identification Toolbox

[dpb]

There are stepwise GLM/LM functions in the Statistics Toolbox that will let one specify an array of possible predictor variables and classes of models that can do a fair amount of the above automagically. It is, of course, still a finite set of models but depending upon the number of variables and the model class, can be a quite large number of possibilities...

There aren't such tools available for nonlinear models, however, so again "it depends" significantly upon just what one is after.