Regression with functions handles-Help please!

Say that

y = a1*f1(x)+ a2* f2(x) + ...+ an*fn(x)

where x and y are column vectors, f1(x),....,fn(x) are function handles at x. How would you solve for the coefficient matrix A using least squares regression?

Here's what I have so far:

function A= MyRegressor(x2,y2,fhs)

for i = 1:numel(fhs)

   f = fhs{i}(x2(i));

end

A=f/y2(i)

Answer 1

[Sean de Wolski]

doc lsqnonlin

Would probably be my first choice.

[SB]

Thanks, I was trying to solve it using just a A=x\y format, however. Is there any way to edit my code to do that?

[SB]

Also, I don't think i'm allowed to use that function.