In the Eq. (10), b1 is set to a column vector of 0.8326/ SPREAD. The SPREAD is the distance between an input vector and a neuron’s weight vector. The data from the input neurons are used to compute an output ai by a typical pattern neuron i, where IWj,i is a specific training vector represented by pattern neuron i, and rj is the smoothing parameter. The shape of the basic function figure in the latent layer No. j is decided by rj. The neurons of the third layer, namely the summation neurons, receive the outputs of the pattern neurons. In the third layer, the outputs from all pattern neurons are augmented. Simple and weighted summations are conducted in neurons of the third layer. The outputting of the network can be represented as Eq. (11). LWk,j is the pattern neuron j connected to weights of the third layer, ai would be obtained in Eq