Method of least squares in combination with Exponential smoothing

26 Views Asked by At

I have been given a formula to compute the squared error of smoothed sequences with the help of the "method of least squares".

There is a sigma $(S_i-X_i)^2$. I have been told that sigma means sum. Does this mean that i need to sum the squared value of (the difference between the smoothed sequence minus demand) of each available row and then take the root?

Example:

Demand --- Smoothed sequence
$165$ --- $163$
$171$ --- $164$

Squared error

$\sqrt{(((163-165)^2 + (164-171)^2)/2)}$

Many thanks in advance.