Self-configuring radial basis function neural networks for chemical pattern recognition

Construction of radial basis function neural networks (RBFN) involves selec tion of radial basis function centroid, radius (width or scale), and number of radial basis function (RBF) units in the hidden layer. The K-means clus tering algorithm is frequently used for selection of centroids and radii. H owever, with the K-means clustering algorithm, the number of RBF units is u sually arbitrarily selected, which may lead to suboptimal performance of th e neural network model. Besides, class membership and the related probabili ty distribution are not considered. Linear averaging (L-A) was devised for selection of centroids and radii for the RBFs and computing the number of R BF units. The proposed method considers the class membership and localized probability density distribution of each class in the training sets. The pa rameters related to the network construction were investigated. The network was trained with the QuickProp algorithm (QP) or Singular Value Decomposit ion (SVD) algorithm and evaluated with the poly(chlorobiphenyl) (PCB) mass spectra and an Italian olive oil reference data set. The prediction accurac y of PCB data sets was better than 94%, and the prediction accuracy with It alian olive oil data sets achieved 100% with RMSEP as low as 2.4 x 10(-3). The training times were usually about a second on a personal computer. The performance of neural networks constructed from the linear-averaging method was observed to be better than that with the K-means algorithm with these data sets.