Journal of Chemometrics

Cover image for Vol. 28 Issue 5

Special Issue: Kowalski Special Issue

May 2014

Volume 28, Issue 5

Pages i–iii, 319–489

Issue edited by: Paul Gemperline

  1. Issue Information

    1. Top of page
    2. Issue Information
    3. Editorial
    4. Special Issue Articles
    1. Issue Information (pages i–iii)

      Article first published online: 9 MAY 2014 | DOI: 10.1002/cem.2541

  2. Editorial

    1. Top of page
    2. Issue Information
    3. Editorial
    4. Special Issue Articles
    1. You have free access to this content
  3. Special Issue Articles

    1. Top of page
    2. Issue Information
    3. Editorial
    4. Special Issue Articles
    1. Process analytical chemistry and chemometrics, Bruce Kowalski's legacy at The Dow Chemical Company (pages 321–331)

      Randy J. Pell, Mary Beth Seasholtz, Kenneth R. Beebe and Mel V. Koch

      Article first published online: 16 SEP 2013 | DOI: 10.1002/cem.2535

      With the passing of Bruce R. Kowalski in 2012, a true visionary for chemometrics and process analytical chemistry has been lost. Bruce made significant contributions in the area of chemometrics and process analytical chemistry when he was a professor of chemistry at the University of Washington. One of the companies that benefitted greatly from Bruce's work was The Dow Chemical Company. This publication attempts to summarize Bruce's legacy at The Dow Chemical Company.

    2. A chemometrics toolbox based on projections and latent variables (pages 332–346)

      Lennart Eriksson, Johan Trygg and Svante Wold

      Article first published online: 13 JAN 2014 | DOI: 10.1002/cem.2581

      A personal view is given about the gradual development of projection methods and their use in chemometrics. From its start around 1970, this development was strongly influenced by Bruce Kowalski and his group in Seattle, and led to the adoption of what in statistics is called the data analytical approach. This approach combined with principal components analysis and later partial least squares worked very well in the analysis of chemical data, and this development is informally summarized by examples and anecdotes.

    3. Characterizing multivariate calibration tradeoffs (bias, variance, selectivity, and sensitivity) to select model tuning parameters (pages 347–357)

      John H. Kalivas and Jon Palmer

      Article first published online: 6 OCT 2013 | DOI: 10.1002/cem.2555

      A new model merit is presented for automatic selection of appropriate tuning parameter values with a good bias/variance balance. The bias/variance balance is related to the underlying model selectivity/sensitivity tradeoff. A new definition of model selectivity is proposed that is able to discern when a model vector deviates from the ideal orthogonal net analyte signal model vector as the well as the degree of deviation. A new merit is also proposed that simultaneously evaluates model selectivity and sensitivity.

    4. Bayesian estimation of membership uncertainty in model-based clustering (pages 358–369)

      Liyuan Chen and Steven D. Brown

      Article first published online: 4 JUL 2013 | DOI: 10.1002/cem.2511

      We illustrate the use of model-based clustering and show three examples in which model-based clustering gives much better performance for overlapping clusters, a more reliable determination of the number of clusters in data, and better identification of clustering than other clustering methods. We also show that Markov chain Monte Carlo simulation, as implemented via Gibbs sampling coupled with model-based clustering, may be used to assess uncertainty of cluster group memberships.

    5. Re-centered kurtosis as a projection pursuit index for multivariate data analysis (pages 370–384)

      Siyuan Hou and Peter D. Wentzell

      Article first published online: 18 NOV 2013 | DOI: 10.1002/cem.2568

      Projection pursuit is a powerful tool for exploratory data analysis and is based on the optimization of a projection index to discover interesting projections of the data. When kurtosis is used as the projection index, unbalanced data sets can lead to poor cluster discrimination in the resulting projections. In this work, a modified algorithm for overcoming this weakness is described and evaluated.

    6. You have free access to this content
      Search prefilters for mid-infrared absorbance spectra of clear coat automotive paint smears using stacked and linear classifiers (pages 385–394)

      Barry K. Lavine, Ayuba Fasasi, Nikhil Mirjankar, Mark Sandercock and Steven D. Brown

      Article first published online: 2 FEB 2014 | DOI: 10.1002/cem.2598

      IR search prefilters for spectral library matching have been developed for the PDQ database using pattern recognition techniques. The importance of feature selection is demonstrated. Identifying specific wavelengths as opposed to informative spectral intervals proved to be a better data analysis strategy.

    7. Multivariate variographic versus bilinear data modeling (pages 395–410)

      Pentti Minkkinen and Kim Harry Esbensen

      Article first published online: 2 AUG 2013 | DOI: 10.1002/cem.2514

      Two contrasting multivariate autocorrelated data sets (a process data series vs. a 1-D geochemical soil profile) are analyzed to illustrate the benefits of using bilinear projection scores for variographic characterization instead of using individual variables. Variograms on a validated number of component scores make a combined multivariate chemometrics – variographic characterization of heterogeneous processes and materials, as well as 1-D transects from stationary objects possible. The usefulness and information obtained in variographic modeling based on scores is illustrated.

    8. Multivariate curve resolution for understanding complex reactions (pages 411–419)

      Randy J. Pell and Xiaoyun Chen

      Article first published online: 9 JUN 2013 | DOI: 10.1002/cem.2507

      Chemometrics is applied to in situ infrared spectra collected from a complex reacting mixture to elucidate the reaction mechanism. A series of models beginning with simple peak area progressing to classical least squares then to multivariate curve resolution with nonnegative spectral and concentration profile constraints and finally to multivariate curve resolution with nonnegative constraints and spectral and concentration profile equality constraints are used. A total of 11 components are used to accurately describe the reacting system.

    9. Comprehensive kinetic model for the dissolution, reaction, and crystallization processes involved in the synthesis of aspirin (pages 420–428)

      David E. Joiner, Julien Billeter, Mary Ellen P. McNally, Ron M. Hoffman and Paul J. Gemperline

      Article first published online: 17 FEB 2014 | DOI: 10.1002/cem.2605

      This paper demonstrates a comprehensive kinetic model for monitoring a batch reaction process in a slurry by in situ attenuated total reflectance ultraviolet–visible spectroscopy. The model includes rates of reaction, dissolution, temperature-dependent solubility, and unseeded crystallization driven by a cooling step. The model accurately estimates time-dependent concentration profiles of dissolved and undissolved components in the slurry and, thereby, the degree of undersaturation and supersaturation necessary for estimation of the rates of dissolution and crystallization. Results were validated across two batches and offline high-performance liquid chromatography measurements.

    10. Concept and role of extreme objects in PCA/SIMCA (pages 429–438)

      Alexey L. Pomerantsev and Oxana Ye Rodionova

      Article first published online: 23 MAY 2013 | DOI: 10.1002/cem.2506

      A novel, semi-robust, data driven technique (DD-SIMCA) is proposed in the PCA/SIMCA context. DD-SIMCA is a dual method of estimation: classical for regular data, and robust for contaminated data. The method provides a clear association with extreme and outlier significance levels. It is shown that being combined with new diagnostic tool called Extreme plot, DD-SIMCA demonstrates a good performance in comparison with the ROBPCA method in the analysis of both regular and contaminated data sets.

    11. Recursive weighted partial least squares (rPLS): an efficient variable selection method using PLS (pages 439–447)

      Åsmund Rinnan, Martin Andersson, Carsten Ridder and Søren Balling Engelsen

      Article first published online: 19 DEC 2013 | DOI: 10.1002/cem.2582

      An important step in fine tuning partial least squares (PLS) regression models is variable selection. In this study, a novel variable selection method named recursive weighted PLS—rPLS—is introduced. It furthermore is compared with two other existing variable selection methods, which it outperforms. rPLS show a good potential as an efficient variable selection method because it is simple and fast and requires little or no user interference.

    12. Impacts of nutrient competition on microalgae biomass production (pages 448–461)

      Lauren H. White, David W. Martin, Khendl K. Witt and Frank Vogt

      Article first published online: 10 SEP 2013 | DOI: 10.1002/cem.2534

      This study investigates impacts of competition among microalgae species for inorganic nutrients, namely bicarbonate, nitrate, and ammonium. It was found that nutrient competition has a considerable influence on the microalgae culture's growth rates, maximum sustainable cell concentration, and cell size distributions. This is of relevance for improving the understanding of the microalgae's role in balancing the atmosphere CO2 concentration by means of sequestering of inorganic nutrients, namely bicarbonate, into biomass.

    13. Multisynchro: a novel approach for batch synchronization in scenarios of multiple asynchronisms (pages 462–475)

      José M. González-Martínez, Onno E. de Noord and Alberto Ferrer

      Article first published online: 24 APR 2014 | DOI: 10.1002/cem.2620

      A novel approach named multisynchro is proposed to overcome the synchronization pitfalls in scenarios of multiple asynchronisms in batch processes. The different types of asynchronisms are effectively detected by using the warping information derived from synchronization. Each set of batch trajectories is synchronized by appropriate synchronization procedures, which are automatically selected based on the nature of asynchronisms present in data. The novel approach also includes a procedure that performs abnormality detection and batch synchronization in an iterative manner.

    14. Recent developments of chemical multiway calibration methodologies with second-order or higher-order advantages (pages 476–489)

      Hai-Long Wu, Yong Li and Ru-Qin Yu

      Article first published online: 22 NOV 2013 | DOI: 10.1002/cem.2570

      This paper reviews chemical multiway calibration methodologies, in particular those with “second-order advantage”, a term coined by Kowalski and co-workers. Among others, some ground-breaking contributions from Kowalski and co-workers in the field of second-order calibration are outlined. Recent developments in multiway calibration, especially in three-way and four-way calibrations, possessing second-order or higher-order advantage with a focus on novel theories and methodologies, as well as practical applications in the quantitative analysis of complicated systems, are also reviewed.