Diagnostic Least Squares
Abstract
In the determination of a set of parameters from a set of observations by the method of least squares, it often occurs that the relationships provided by the observations are not really linearly independent when the random errors in the observations are considered. A method for finding physically reasonable parameters and confidence limits for parameters is described. This method is based on parameter scaling and diagonalization of the matrix of the normal equations.
- Publication:
-
Journal of Computational Physics
- Pub Date:
- December 1970
- DOI:
- 10.1016/0021-9991(70)90036-7
- Bibcode:
- 1970JCoPh...6..367C