Qualitative Analysis for 1- D Recurrent Neural Networks
SU Jua · Journal of Chengdu Normal University · 2015
This paper studies the qualitative characteristics of 1- D recurrent neural networks. By employing the formula of cubic function root,the relationships between the coefficients of the system and the number of equilibria are revealed. Furthermore,the stability and attractive domains of all equilibria are derived by using the Taylor theorem and the monotonicity. This paper improves and completes the previous conclusions. Finally,two examples and simulations are illustrated to validate our theory.