Using fminsearch to minimize root mean square error.

_After searching around the net for a while I have not found a good answer to this question. Please point me to one or help out if possible.

I have created a function to calculate the root mean square error of a column vector K with a scalar x:

function  [rootmeansquare] = RMSE (K, x)
rootmeansquare = sqrt ( sum ( (K - x) .^ 2 )

I want to minimize RMSE for a defined vector K by adjusting x. Can I use the fminsearch function to do this? If so how?

Answer 1

[Paulo Silva]

K=[1 2 3 4 5 6]; %some data to test
rms =@(x)sqrt(sum((K - x).^ 2)); %function similar to yours
fminsearch(rms,0) %find the x

Another way

x=-20:0.01:20;
y=zeros(1,numel(x));
for i=1:numel(x)
y(i)=sqrt(sum((K - x(i)).^ 2));
end
ymin=min(y)
xmin=x(y==ymin)
hold on
plot(x,y)
plot(xmin,ymin,'ro')
text(xmin,ymin,['xmin=' num2str(xmin) char(10) 'ymin=' num2str(ymin)])