Iv. Tetko et al., Polynomial neural network for linear and non-linear model selection in quantitative-structure activity relationship studies on the internet, SAR QSAR EN, 11(3-4), 2000, pp. 263-280
This article presents a self-organising multilayered iterative algorithm th
at provides linear and non-linear polynomial regression models thus allowin
g the user to control the number and the power of the terms in the models.
The accuracy of the algorithm is compared to the partial least squares (PLS
) algorithm using fourteen data sets in quantitative-structure activity rel
ationship studies. The calculated data show that the proposed method is abl
e to select simple models characterized by a high prediction ability and th
us provides a considerable interest in quantitative-structure activity rela
tionship studies. The software is developed using client-server protocol (J
ava and C++ languages) and is available for world-wide users on the Web sit
e of the authors.