ACD/Labs 5^th Annual European Users' Meeting 2004 (EUM)

Advanced UV-IR and Curve Processing

Michael Boruta

Abstract

Calibration and prediction of data is often a non-trivial task in spectroscopy. The ability to pick a well behaved set of data can be a time consuming process, with the selection of the calibration region having an effect on the success of the prediction ability of the model.

With much of the success of the model based on the users' ability to select a good frequency upon which to base the calibration, we decided to look at the ability of the SIMPLISMA algorithm to help us pick a frequency to base the calibration on. This presentation discusses why SIMPLISMA was chosen and looks at the predictive ability of a set of calibration data viewed from several different methods.

Download the presentation in MS PowerPoint (512 Kb ZIP file) or Adobe Acrobat format (514 Kb PDF file).

Relevant Products: UV-IR Processor