I have a set of data-points with their measurement errors, and I'm usinga least mean square method to fit a curve to them (minimizing the sum of squared distances to the curve). What I want to know is how do you calculate errors on the estimated parameters of the fitted curve? I know the usual error propagation rules for sums and products, but what do you do with the minimizing bit?
C. . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
