Title :
A theorem of equivalence on the methods of least-squares estimation
Author_Institution :
California Institute of Technology, Pasadena, CA, USA
fDate :
2/1/1985 12:00:00 AM
Abstract :
A theorem on the methods of least-squares estimation is stated and proved. This theorem enables the replacement of a system of correlated measurements by an equivalent system of uncorrelated measurements without a whitening process, thus simplifying the analysis while resulting in the same minimum-variance estimate.
Keywords :
Least-squares methods; Covariance matrix; Eigenvalues and eigenfunctions; Error correction; Estimation error; Kalman filters; Matrices; Matrix decomposition; Noise measurement; Signal to noise ratio; Vectors;
Journal_Title :
Automatic Control, IEEE Transactions on
DOI :
10.1109/TAC.1985.1103911
