GA codes for linear regression equation

How can i code an optimization GA code for a multiple regression equation in the form Y=Ax1 +BX2 + C;

where X1 and X2 are variables and A,B,C are the constants for optimization.

Thanks in advance

Answer 1

[Abdolkarim Mohammadi]

Although ga() can fit multiple linear regression models, it is recommended to use regress() since it is dedicated to linear regression and is faster and more accurate than ga(). By the way, you can get this code from here:

https://www.mathworks.com/matlabcentral/answers/567840-genetic-algorithm-to-optimize-the-variable-of-linear-regression-a-b1-b2#answer_468399

Answer 2

[Star Strider]

Since the fitness function must return a scalar value to the ga function, I would do something like this:

x = [x1(:) x2(:)];                                  % Matrix Of Column Vectors
y = y(:);                                           % Column Vector
model = @(b,x) b(1).*x(:,1) + b(2).*x(:,2) + b(3);  % Define Linear Regression Model
ftns = @(b) norm(y - model(b,x));                   % Fitness Function

The ga function would then return the optimised values for the ‘b’ parameters. This approach can be used with any regression equation.

I have not tested this function specifically, however it should work.