Linear weighting
Posted: Mon May 15, 2006 4:08 pm
Hi everybody
We have more options in MS for regression purpose. Linear 1/x,Linear1/x2
Quadratic 1/y and Quadratic 1/y2. Generally all are preferring with linear 1/x2 . Why? How to select the regression for calibration curve? Any specific method is there? I have read some of other compnies document, they have given some table for why they have selected Quadratic.
But i cant understand.
So please help me.[/quote]