Journal
ANALYTICA CHIMICA ACTA
Volume 1058, Issue -, Pages 9-17Publisher
ELSEVIER SCIENCE BV
DOI: 10.1016/j.aca.2018.10.055
Keywords
Signal pre-processing; Pre-treatment; Measurement units; Derivatives; Standard normal variate (SNV); Loadings
Categories
Funding
- Italian Ministry of Education, Universities and Research (MIUR) [RBSI14CJHJ, CUP: D32I15000150008]
Ask authors/readers for more resources
The present tutorial paper is aimed to analyse and critically discuss the consequences of row pre-processing (conversion of measurement units, derivatives, and standard normal variate transform) on the evaluation of final outcomes of chemometric data analysis. An in-depth focus on pre-processing effects both on the signal shape and on misinterpretation of results - a crucial and disregarded issue in the analytical field - is presented. It is shown how this preliminary step of data processing may lead, in many cases, to draw incongruous conclusions, not actually based on real information embodied within data, but on artefacts arising from the mathematical transforms. This tutorial is not limited to a description of the problem, it also introduces strategies and tools for overcoming such unwanted effects, allowing a direct interpretation of the importance of original variables to be performed, explaining the chemical information that actually characterises samples. The dangerous implications of row preprocessing on instrumental signals is demonstrated on real datasets coming from different analytical techniques: transmission and attenuated total reflection infrared spectroscopy, cyclic voltammetry, X-ray fluorescence spectroscopy, Raman spectroscopy, and ultraviolet-visible spectroscopy. Hence, the impact of this widespread problem in most of the branches of analytical chemistry is illustrated. (c) 2018 Elsevier B.V. All rights reserved.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available