Abstract
The lower bounds for the a posteriori prediction error of a nonlinear predictor realized as a neural network are provided. These are obtained for a priori adaptation and a posteriori error networks with sigmoid nonlinearities trained by gradient-descent learning algorithms. A contractivity condition is imposed on a nonlinear activation function of a neuron so that the a posteriori prediction error is smaller in magnitude than the corresponding a priori one. Furthermore, an upper bound is imposed on the learning rate ? so that the approach is feasible. The analysis is undertaken for both feedforward and recurrent nonlinear predictors realized as neural networks.
Original language | English |
---|---|
Pages (from-to) | 1285-1292 |
Number of pages | 8 |
Journal | Neural Computation |
Volume | 12 |
Issue number | 6 |
DOIs | |
Publication status | Published - 1999 |