Title :
On index sum of continuous time neural network
Author :
Liu, Tundong ; Yang, Xiao-song
Author_Institution :
Dept. of Autom., Xiamen Univ., Xiamen, China
Abstract :
In this paper a theory of index sum of equilibrium points of neural networks described by ordinary differential equations is presented. First a framework for index sum of equilibrium points of bounded dynamical systems is established, then a formula of index sum of equilibrium points of neural networks described by ordinary differential equations is discussed by virtue of the established framework, and an estimate of the number of asymptotically stable equilibrium points in a neural network is also presented.
Keywords :
continuous time systems; differential equations; neural nets; nonlinear dynamical systems; bounded dynamical systems; continuous time neural network; equilibrium points; index sum; ordinary differential equations; Artificial neural networks; Differential equations; Eigenvalues and eigenfunctions; Equations; Indexes; Jacobian matrices; Trajectory; bounded dynamical systems; equilibrium points; neural networks;
Conference_Titel :
Natural Computation (ICNC), 2010 Sixth International Conference on
Conference_Location :
Yantai, Shandong
Print_ISBN :
978-1-4244-5958-2
DOI :
10.1109/ICNC.2010.5583213
