Title :
Dynamical neural network for solving a class of nonlinear equation systems
Author :
Xia, You-shen ; Ye, Da-zhen
Author_Institution :
Dept. of Math., Nanjing Univ. of Posts & Telecommun., China
Abstract :
This paper presents a dynamical neural network for solving a class of nonlinear equation systems, and shows that it globally converges to an exact solution. In hardware implementations, the network architecture does not contain the complex structure which implements the Jacobian matrix. Finally, an application of the neural network approach to two-point boundary value problems is given
Keywords :
nonlinear equations; Jacobian matrix; complex structure; differential equations; dynamical neural network; global convergence; neural network; nonlinear equation systems; two-point boundary value problems; Boundary value problems; Computer architecture; Computer networks; Differential equations; Jacobian matrices; Neural network hardware; Neural networks; Newton method; Nonlinear equations;
Conference_Titel :
Neural Networks, 1996., IEEE International Conference on
Print_ISBN :
0-7803-3210-5
DOI :
10.1109/ICNN.1996.549178
