How to estimate weights of the weighted mean?

1 view (last 30 days)
Anudeep
Anudeep on 25 Nov 2020
Edited: Jason Nicholson on 29 Dec 2020
I have an equation of y = (a1x1 + a2x2 + ... + a81x81)/81
I have a set of y and x values to estimate the coefficients a1, a2, ... , a81.
How do I code to derive the coefficients?
  1 Comment
KSSV
KSSV on 25 Nov 2020
Edited: KSSV on 25 Nov 2020
Read about regression, '\', linear least squares.

Sign in to comment.

Answers (1)

Jason Nicholson
Jason Nicholson on 29 Dec 2020
Edited: Jason Nicholson on 29 Dec 2020
Put all the each x and y pair on a row. Formulate the matrix A containing all the x1...x81 and x values. Put the y values in a column matrix. Use backslash to linear least squares solve.
pseudo code:
assume each column of x is x1, x2, etc and each row is a point.
A = [x(:,1) x(:,2) ... x(:,81)] in other words, A = x
a = A\y
Note, there is a lot of issues related to stability of this problem. Research linear least squares problem. This is a good place to start: Systems of Linear Equations
Please, remember to accept my answer if it answers your question.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!