Publication Details

Category Text Publication
Reference Category Journals
Title (Primary) Influence of data preprocessing and kernel selection on probabilistic neural network modeling of the acute toxicity of chemicals to the fathead minnow and Vibrio fischeri bacteria
Author Niculescu, S.P.; Kaiser, K.L.E.; Schüürmann, G.
Source Titel Water Quality Research Journal of Canada
Year 1998
Department OEC; COE
Volume 33
Issue 1
Page From 153
Page To 165
Language englisch
Abstract

We investigated the connection between the data preprocessing strategy and kernel choice on the quality of the associated basic probabilistic neural network models for the acute toxicity of various chemicals to the fathead minnow and to Vibrio fischeri bacteria. The models employ exclusively structural parameters and physicochemical properties as inputs. Results show that the Gaussian kernel is preferable overthe reciprocal kernel model. Data preprocessing based on the hyperbolic tangent and the sigmoid logistic transforms provides the best results at the level of the cross validation experiment. Improved modelsbased on cross validation partial models and linear corrections werealso investigated. The results show that the improved models with data preprocessing based on the hyperbolic tangent and the finite interval transforms are the best with practically identical quality of predictions.

Persistent UFZ Identifier https://www.ufz.de/index.php?en=20939&ufzPublicationIdentifier=8887
Niculescu, S.P., Kaiser, K.L.E., Schüürmann, G. (1998):
Influence of data preprocessing and kernel selection on probabilistic neural network modeling of the acute toxicity of chemicals to the fathead minnow and Vibrio fischeri bacteria
Water Qual. Res. J. Canada 33 (1), 153 - 165