Classification: Basic Concepts

B. K. Lavine, W. S. Rayens

Research output: Chapter in Book/Report/Conference proceedingChapterpeer-review

18 Scopus citations

Abstract

Classification of data is an important subject in chemometrics as evidenced by the large number of citations appearing in the Chemical Abstract database annually on supervised pattern recognition techniques. Developing a classifier from spectral or chromatographic data is desirable for any number of reasons including source identification, food quality testing, detection of odorants, presence of disease in a patient or animal from which a sample has been taken, and the detection of a specific analyte to name a few. However, in chemistry the scope of the analysis will often go beyond a simple classification of the data. Higher levels of classification for chemical data include but are not limited to detection of an unknown class in the training or validations set, the importance of individual measurement variables to the overall classification of the data, and the characterization of the internal class structure, for example, the presence of the asymmetric data structure in the training set.

Original languageEnglish
Title of host publicationComprehensive Chemometrics
Pages507-515
Number of pages9
Volume3
DOIs
StatePublished - 2009

Keywords

  • Artificial neural networks
  • Backpropagation neural networks
  • Classification
  • Classification trees
  • Counterpropagation neural networks
  • Decision trees
  • Regularized discriminants
  • SIMCA
  • Supervised pattern recognition techniques
  • Support vector machines

ASJC Scopus subject areas

  • Biochemistry, Genetics and Molecular Biology (all)

Fingerprint

Dive into the research topics of 'Classification: Basic Concepts'. Together they form a unique fingerprint.

Cite this