Author |
Message |
Erik Skibsted
| Posted on Thursday, August 02, 2001 - 5:02 am: | |
Dear Fellow NIR collegues I am currently working with quantitative modelling of constituents in powder mixtures. I have made some initial trials using net analyte signal (NAS) calibration plots and I am now ready for a "bigger" experiment where I want to compare the performance of NAS vs. PLS. I would to hear if anybody in this forum has experience with NAS and meaby point me in the direction of some literature. I have some from Farber, Wentzell and Lorber. |
hlmark
| Posted on Thursday, August 02, 2001 - 6:49 am: | |
Erik - you might want to try posting your question to the Chemometrics Discussion group. You can get info the group by sending a message to: International Chemometrics Society at web address: [email protected] You'll probably get a message back that you need to join the discussion group, but I think that message will tell you how to do that. Unfortunatly I don't seem to have that info handy. If you have a problem, let me know & I'll look it up. Howard |
Christopher D. Brown
| Posted on Thursday, August 02, 2001 - 9:16 am: | |
Erik, The NAS is estimated by every linear calibration method, PLS included. By NAS calibration, then, are you referring to an estimated NAS by direct knowledge of your calibration mixtures (i.e., known pure-component spectra, CLS-style)? This is in the flavor of one of Ave Lorber's papers (and this may be the one you have on your desk). Perhaps you could elaborate a bit more, and as Howard suggested, cross-post your message to the ICS-listserver. Chris |
|