It is shown in this paper that if the hidden layer units take a sinusoidalactivation function,the optimum weights of the three-layer feedforward neural networkc
The method of automatically building 3-D molecular structure ofmacrorings is proposed in the paper.The method can build the 3-D models ofmacrorings full automat
Two diagnostics are presented for the transformation of explanatoryvariables in regression.One is based on the likelihood displacement proposed by Cookand Weisb
In this paper,the problem of Parameter estimation in linear delayedsystems from sampled data is treated.Using numerical integral operation(NIO),anidentification