The use of the Durbin-Watson criterion for noise and background reduction of complex liquid chromatography/mass spectrometry data and a new algorithm to determine sample differences

Recently, Massart's research group showed the potential of the Durbin-Watson criterion as a tool to determine the pseudo-rank of mixture data sets using stepwise interactive self-modeling mixture analysis approaches. The Durbin-Watson criterion is a simple and intuitive equation which works surprisingly well for other data analysis problems. As an example, the use of the Durbin-Watson criterion will be introduced in this paper as a tool to reduce noise and baseline problems in liquid chromatography/mass spectrometry (LC/MS) by variable reduction. In addition, another algorithm to extract differences between highly related samples will be presented. Both new approaches have some advantages above previously published methods. (c) 2004 Elsevier B.V. All rights reserved.