Multiple shape basis function networks for rule based analysis of data

The concept of radial basis functional networks has been generalised by introducing multiple shape basis function networks. The learning algorithm, restricted Coulomb energy learning, capable of generating the hidden layer dynamically, is extended to include new components to adjust the shape of the region of attraction of a prototype neuron, in addition to the adaptation of centre weight and radial parameters so that the input space is covered more efficiently by automatically generating clusters of different sizes and shape. This shows a clear reduction in the number of neurons or the number of fuzzy rules generated and the classification accuracy is increased significantly. This improvement is highly relevant in developing neural networks which are functionally equivalent to fuzzy classifiers since the transparency is strongly related to the compactness of the generated system