Journal of Chemometrics

Cover image for Vol. 25 Issue 5

Special Issue: 5th International Symposium on Computer Applications and Chemometrics in Analytical Chemistry, Budapest, Hungary, June 21–25, 2010. Guest Editor: Károly Héberger

May 2011

Volume 25, Issue 5

Pages 225–286

Issue edited by: Karoly Heberger

  1. Special Issue Articles

    1. Top of page
    2. Special Issue Articles
    1. One-class classifiers (pages 225–246)

      Richard G. Brereton

      Version of Record online: 24 MAY 2011 | DOI: 10.1002/cem.1397

      The main ideas of one-class classifiers are introduced. They are illustrated using mainly quadratic discriminant analysis and support vector domain description, plus disjoint and conjoint principal component models. Examples are drawn from metabolomics, thermal analysis and simulations.

    2. Destruction of normal distribution in small samples by centering and scaling (pages 247–253)

      Gergely Tóth

      Version of Record online: 4 MAR 2011 | DOI: 10.1002/cem.1382

      We show in our comparative study that centering and especially scaling in small samples of normally distributed data require consideration as a first step in data evaluation. Depending on the constants applied in centering and scaling the transformed variables retain the nature of the original distribution, or they follow well-known distributions as the t-distribution, but in some cases bimodal, Dirac-delta, Viking helmet like and uniform distributions appear.

    3. Hasse diagram technique as exploratory tool in sediment pollution assessment (pages 254–261)

      Stefan Tsakovski and Vasil Simeonov

      Version of Record online: 4 MAR 2011 | DOI: 10.1002/cem.1381

      An original strategy for sediment pollution assessment consisting of Hasse diagram technique and SOM as pre-treatment technique was used. The advantages of the SOM algorithm as proper classification of environmental data sets and decision support abilities of Hasse diagram technique were applied to establish the relationship between the selected chemical pollutants and the sampling locations in order to distinguish between specific reasons for sediment pollution.

    4. Chemometric assessment of the semivolatile organic contaminants content in the atmosphere of the selected sites in the Republic of Macedonia (pages 262–274)

      Trajče Stafilov, Biljana Škrbić, Jana Klánová, Pavel Čupr, Ivan Holoubek, Marin Kočov and Nataša Đurišić-Mladenović

      Version of Record online: 17 MAR 2011 | DOI: 10.1002/cem.1374

      The levels of 12 organochlorine pesticides, 7 indicator polychlorinated biphenyls, and 28 polycyclic aromatic hydrocarbons in the air of six locations in the Republic of Macedonia have been interpreted by principal component and cluster analysis. The obtained results gave important information about the potential polluting sources in the regions of interest, indicating some specific emissions. Influence of the available meteorological parameters on the SVOC atmospheric concentrations have been evaluated by multiple regression analysis.

    5. Comparison of novel sensory panel performance evaluation techniques with e-nose analysis integration (pages 275–286)

      László Sipos, Zoltán Kovács, Dániel Szöllősi, Zoltán Kókai, István Dalmadi and András Fekete

      Version of Record online: 30 MAR 2011 | DOI: 10.1002/cem.1391

      A complex approach was used to evaluate the performance of a sensory panel with two novel techniques: GCAP (Gravity Center Area/Perimeter) and CRRN (Compare Ranks with Random Numbers). Profile analysis was performed on Sri Lanka black tea batches from different plantations; for the prediction of sensory data from electronic nose results partial least square regression and support vector machine regression were used. Prediction by support vector machine gave close correlation between the results of electronic nose measurement and odor attributes.