Error estimation of recurrent neural network models trained on a finite set of initial values

Addresses the problem of estimating training error bounds of state and output trajectories for a class of recurrent neural networks as models of nonlinear dynamic systems. The bounds are obtained provided that the models have been trained on N trajectories with N independent random initial values which are uniformly distributed over [a, b]/sup m//spl epsiv//spl Rscr//sup m/.