Write Objective function with known linear regression model

I am studying the tutorial on this page where a linear regression model is obtained and needs to be optimised.

https://uk.mathworks.com/help/stats/improve-an-engine-cooling-fan-using-design-for-six-sigma-techniques.html

In this tutorial I see this step:

Write the objective function.

f = @(x) -x2fx(x,'quadratic')*mdl.Coefficients.Estimate;

I wonder how -x2fx(x,'quadratic') is obtained in this example? I did not see it from the linear model

Cheers

Answer 1

[Sean de Wolski]

This is kind of the old way of doing it. I would just use the predict method of the object to make the prediction.

f = @(x) -predict(mdl, x);

Feel free to email me, I have an updated copy of this example.

[Pingfei Jiang]

Hi Sean, is it possible for you to email me the updated example? I am still struggle to understand this example. Many thanks.