EVALUATION OF SIMCA PATTERN RECOGNITION FOR USE WITH POLYCHLORINATED BIPHENYL DATA SETS.

Published

Journal Article

The evaluation of SIMCA (Soft Independent Modeling by Class Analysis) was done by two separate sets of experiments. The first set used SIMCA to formulate principal component models (PC models) of each Aroclor class, such that the models could be used to uniquely classify a sample as containing a particular Aroclor. Experiments in this set evaluated four different data scaling/normalization procedures which were carried our prior to the calculation of PC models. The second set of experiments used SIMCA to examine clustering of the adipose tissue samples and to classify the sample as to predominant Aroclor type.

Duke Authors

Cited Authors

  • Moseley, MA; Pellizzari, ED; Williams, LR

Published Date

  • December 1, 1984

Published In

  • National Meeting American Chemical Society, Division of Environmental Chemistry

Volume / Issue

  • 24 / 2

Start / End Page

  • 63 - 64

International Standard Serial Number (ISSN)

  • 0270-3009

Citation Source

  • Scopus