Journals / Turkish Journal of Mathematics / 2001 / Cilt: 25 - Sayı: 2
Global Convergence for Discrete Dynamical Systems and Forward Neural Networks
- Pages
- 345–354
- DOI
- —
Abstract
Using a theoretical result regarding the global stability of discrete dynamical systems of lower triangular form, we establish convergence properties of forward neural networks when the neuron response functions fail to be continuous.
Özet
Using a theoretical result regarding the global stability of discrete dynamical systems of lower triangular form, we establish convergence properties of forward neural networks when the neuron response functions fail to be continuous.