Journal of Chemometrics v.25 n.3 2011

1. 
   
   Application of EEM fluorescence in combination with PARAFAC analysis to simultaneously monitor quercetin in its deprotonated, aggregated, and protein bound states (pages 101–108)

   John V. Simpson, Marissa Burke and Renee D. Jiji

   Article first published online: 1 SEP 2010 | DOI: 10.1002/cem.1325

   Excitation-emission matrix (EEM) fluorescence in combination with weighted-parallel factor analysis (WPARAFAC) was employed to characterize the pH and concentration dependent fluorescence of aqueous quercetin. Distinct pH and concentration dependent fluorescence profiles were observed. The fluorescence of quercetin was also characterized in the presence of bovine serum albumin, which resulted in a narrowing and red shift and of the excitation maxima. Coupling WPARAFAC and EEM fluorescence, enabled simultaneous detection of multiple fluorescent quercetin species in solution.

2. Partial least squares discrimination with heterogeneous covariance structures(pages 109–115)

   Yushu Liu, William Rayens, Anders Andersen and Charles Smith

   Article first published online: 2 FEB 2011 | DOI: 10.1002/cem.1352

   This paper addresses high-dimensional classification wherein the within-groups covariances are potentially not the same and, hence, a single reduction step, followed by linear discrimination may not be appropriate. A variety of techniques are compared by way of a large-scale simulation study. An application aimed at distinguishing asymptomatic women at risk for Alzheimer's disease from women not at risk is presented.

3. One stop shopping: feature selection, classification and prediction in a single step(pages 116–129)

   Barry K. Lavine, Kadambari Nuguru and Nikhil Mirjankar

   Article first published online: 16 FEB 2011 | DOI: 10.1002/cem.1358

   A genetic algorithm that performs feature selection, classification, and prediction in a single step is reported. The advantages of using this method for pattern recognition are demonstrated in two studies. In one study, Raman spectroscopy and the pattern recognition are used to develop a potential method to discriminate hardwoods, softwoods, and tropical woods. In a second study, biopsy material analyzed by cDNA microarrays is identified as to type.

   • Abstract
   • Full Article (HTML)
   • PDF(879K)
   • References
4. You have full text access to this content
   Stacked PLS for calibration transfer without standards (pages 130–137)

   Wangdong Ni, Steven D. Brown and Ruilin Man

   Article first published online: 1 FEB 2011 | DOI: 10.1002/cem.1369

   The successful application of stacked partial least-squares (SPLS) regression for direct application of multivariate calibration models to data from a secondary spectrometer, without use of any calibration transfer, is reported. Predictive results comparable to those obtained from PLS calibration followed by a calibration transfer done with transfer standards are demonstrated.