Journal of Chemometrics

Cover image for Vol. 26 Issue 8-9

Special Issue: Proceedings of the 12th Scandinavian Symposium on Chemometrics, SSC12, held in June 2011

August-September 2012

Volume 26, Issue 8-9

Pages 423–495

Issue edited by: Åsmund Rinnan

  1. Meeting Report

    1. Top of page
    2. Meeting Report
    3. Special Issue Articles
  2. Special Issue Articles

    1. Top of page
    2. Meeting Report
    3. Special Issue Articles
    1. Estimation of volume fraction and flow regime identification in inclined pipes based on gamma measurements and multivariate calibration (pages 425–434)

      Benjamin Kaku Arvoh, Rainer Hoffmann, Arne Valle and Maths Halstensen

      Article first published online: 1 APR 2012 | DOI: 10.1002/cem.2437

      Multivariate calibration was applied to data obtained from a traversable dual-energy gamma densitometer for volume fraction estimation and flow regime identification. The estimated mixture densities were compared with those obtained from a single-energy gamma densitometer. The results from the estimated volume fractions and the flow regime identification properties of the models were accurate.

    2. Subspace methods for dynamic model estimation in PAT applications (pages 435–441)

      Jonas Hoeg Thygesen and Frans W. J. van den Berg

      Article first published online: 28 MAR 2012 | DOI: 10.1002/cem.2424

      This paper introduces a class of dynamic estimators called the state space models. It illustrates how subspace methods for state space modeling are closely related to known chemometric tools and how state space models can be applied in feed-forward process monitoring. This is carried out by showing how a (nonlinear) milk coagulation process can be approximated by a linear-estimated dynamic model.

    3. Qualitative detection of illegal drugs (cocaine, heroin and MDMA) in seized street samples based on SFS data and ANN: validation of method (pages 442–455)

      Jekaterina Mazina, Valeri Aleksejev, Tatjana Ivkina, Mihkel Kaljurand and Larissa Poryvkina

      Article first published online: 26 JUN 2012 | DOI: 10.1002/cem.2462

      In this paper, the validation procedure of spectral fluorescence signature method combined with multilayer perceptron artificial neural networks for detection of illegal drugs (cocaine, heroin and 3,4-methylenedioxy-N-methylamphetamine) in street samples is proposed. The qualitative information, based on a binary response (detected/not detected), was directly obtained through the response of an expert system. The performance parameters (limit of detection, selectivity/matrix effects, threshold value and robustness) were evaluated according to the requirements for qualitative method.

    4. Validation of model of multivariate calibration: an application to the determination of biodiesel blend levels in diesel by near-infrared spectroscopy (pages 456–461)

      Werickson Fortunato de Carvalho Rocha, Raquel Nogueira and Boniek Gontigo Vaz

      Article first published online: 16 APR 2012 | DOI: 10.1002/cem.2420

      This paper describes the development and validation of a multivariate calibration model on the basis of partial least squares and net analyte signal, which can be directly applied to determine biodiesel concentrations between 2%–90% in biodiesel/diesel blends analyzed by near-infrared spectroscopy. The model was validated, with regard to accuracy, limit of detection, limit of quantification, sensitivity, and selectivity, through the calculation of the corresponding figures of merit.

    5. Use of partial least squares discriminant analysis on visible-near infrared multispectral image data to examine germination ability and germ length in spinach seeds (pages 462–466)

      Nisha Shetty, Merete Halkjær Olesen, René Gislum, Lise Christina Deleuran and Birte Boelt

      Article first published online: 15 MAR 2012 | DOI: 10.1002/cem.1415

      Partial least squares discriminant analysis (PLS-DA) was applied on features extracted from multispectral images of spinach seeds. The PLS-DA prediction resulted in an independent test set not only providing discrimination of seed size but also demonstrating how germination ability and germ length vary according to seed size. The result indicated that larger seeds had both a significantly higher germination potential and germ length compared with smaller seeds. The variable importance for projection method showed that the near infrared wavelength region is important for germination predictability.

    6. On the effectiveness of cross-fitting in multi-block PLS (CF-MBPLS) (pages 467–473)

      Jarno Kohonen, Hannu Alatalo and Satu-Pia Reinikainen

      Article first published online: 1 MAR 2012 | DOI: 10.1002/cem.1413

      Overfitting in multi-block PLS (partial least squares or projection to latent structures) method can be reduced with a recent development of cross-fitting the score values. This can reduce the effect of outliers in the data and is also more sensitive to any nonlinearity in the data. Cross-fitting was adapted to multi-block PLS with automatic selection of latent variables using ANOVA approach and removal of outliers.

    7. Resolution of spectrally rank-deficient multivariate curve resolution: alternating least squares components in comprehensive two-dimensional liquid chromatographic analysis (pages 474–486)

      Christophe Tistaert, Hope P. Bailey, Robert C. Allen, Yvan Vander Heyden and Sarah C. Rutan

      Article first published online: 15 MAR 2012 | DOI: 10.1002/cem.2434

      A new constraint for multivariate curve resolution-alternating least squares (MCR-ALS) is proposed for resolving comprehensive liquid chromatographic peaks appearing within the same MCR-ALS component and dividing them among additionally created components with highly similar spectral characteristics. Although manual quantification results in comparable percent relative standard deviations prior and after application of the constraint, the presence of a single compound in each chromatographic component allows for automated quantification of the compounds by simple summation of the second-dimension chromatograms.

    8. Significance of the structure of data in partial least squares regression predictions involving both natural and human experimental design (pages 487–495)

      Åsmund Rinnan and Lars Munck

      Article first published online: 17 APR 2012 | DOI: 10.1002/cem.2438

      We are aiming at opening a Pandora's box of how the prediction of protein proceeds from near-infrared spectra using partial least squares regression in a unique set of chemically diverse barley mutant samples. An external validation of the sources of co-variation in nature that are exploited by chemometric models would give a framework for manipulating the deciding information to make expensive calibration more economical.