Premium
A new maximum entropy‐based method for deconvolution of spectra with heteroscedastic noise
Author(s) -
Buttingsrud Bård,
Alsberg Bjørn K.
Publication year - 2004
Publication title -
journal of chemometrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.47
H-Index - 92
eISSN - 1099-128X
pISSN - 0886-9383
DOI - 10.1002/cem.898
Subject(s) - deconvolution , heteroscedasticity , computer science , robustness (evolution) , noise (video) , blind deconvolution , algorithm , mathematics , artificial intelligence , machine learning , chemistry , biochemistry , image (mathematics) , gene
Broadening of spectral lines combined with large and heteroscedastic noise contributions constitutes an important problem in analytical chemistry. Reduced interpretability and artefacts in further data analysis make deconvolution methods necessary. A new robust deconvolution method (RHEMEM) based on the principle of maximum entropy is proposed in order to effectively handle the presence of heteroscedastic noise. Other deconvolution methods such as Jansson's method, Fourier self‐deconvolution and LOMEP are also studied with respect to their ability to handle heteroscedastic noise. A systematic simulation study is used to compare the performance of the new method with the reference methods. They are evaluated according to reconstruction performance, robustness and the ability to work without manual input. Copyright © 2005 John Wiley & Sons, Ltd.