On the method of least-squares
WH Jefferys - ASTRONOMICAL JOURNAL V. 85, P. 177, 1980, 1980 - adsabs.harvard.edu
A method for fitting models to observed data by least squares when the data appear
nonlinearly in the equations of condition, where there are nonlinear constraints among
model parameters, and where the observations may be correlated, is described. The method
is independent of the form in which the equations of condition are expressed, and is a
generalization of the classical technique. The method is nearly as simple to apply in practice
as the classical method of least squares.
nonlinearly in the equations of condition, where there are nonlinear constraints among
model parameters, and where the observations may be correlated, is described. The method
is independent of the form in which the equations of condition are expressed, and is a
generalization of the classical technique. The method is nearly as simple to apply in practice
as the classical method of least squares.