MATLAB: Minimize the error between the measured y values when give a set of data:

MATLAB

I'm having trouble understanding this problem and assume that we'll trying to find the best fit curve:

I do have some coding done, but I'm not sure if I'm going in the right direction. I would like if this was explain clearly. I have also attached the given data that contains 2 matfiles x and y_noise. Thanks.

load 'prob3.mat'

c_1 = x;

c_2 = y_noise;

y = c_1.*cos(x) + c_2.*exp(-x)

This is my beginning code. But looks off from what is being asked.

Best Answer

Use an anonymous function to create your ‘y’ function, then use the norm function to create your cost function, then the fminsearch function to do the ‘minimize’ step. See the section on ‘Anonymous Functions’ in Function Basics for details.

If you want to write a fitting function for:

y = m*x + b

you would write it as:

yfun = @(b,x) b(1).*x + b(2);

That should give you a hint on how to go about this.

Related Solutions

MATLAB: OLS estimator b=x\y

‘What is difference between these two expressions?’

The way they are implemented in MATLAB. The documentation for mldivide,\ explains how it works (so I won’t go into that here), and the reason it is more stable and preferred over taking the inverse and multiplying.

MATLAB: Reaching the element of an array passed to a function

The first step needs to be to find the indices of the array that correspond to the conditional statement. You can use something like find for this, but logical indexing is usually preferred.

y = ones(size(A)); % Set up y.
ind = A >= C; % First, find elements of A that are greater than C.
y(ind) = 0;
y(~ind) = 1;

The last line is not absolutely necessary in this case, since we initialized y using ones, but you could put an equation here if something more complicated was needed.

Best Answer

Related Solutions

MATLAB: OLS estimator b=x\y

MATLAB: Reaching the element of an array passed to a function

Related Question