Jc. Olivier, OBTAINING OPTIMUM MINIMUM POINTS OF ERROR FUNCTIONS - A NEURAL-NETWORK APPROACH, IEEE transactions on circuits and systems. 2, Analog and digital signal processing, 42(11), 1995, pp. 722-725
An iterative neural network solution for the problem of obtaining the
optimal minima of error or cost functions containing many local minima
is proposed, Use is made of an adaptive constraint function eliminati
ng local minima previously encountered, It is shown that the method yi
elds very fast response times, and the method is applied to a represen
tative example with simulation results verifying its usefulness and sp
eed.