Analog VLSI and neural systems
Analog VLSI and neural systems
Journal of Computational and Applied Mathematics - Special issue on numerical analysis 2000 Vol. IV: optimization and nonlinear equations
Numerical Methods Using MATLAB
Numerical Methods Using MATLAB
Construction of Newton-like iteration methods for solving nonlinear equations
Numerische Mathematik
ISICA '08 Proceedings of the 3rd International Symposium on Advances in Computation and Intelligence
Expert Systems with Applications: An International Journal
ICICS'09 Proceedings of the 7th international conference on Information, communications and signal processing
IEEE Transactions on Neural Networks
A recurrent neural network for solving Sylvester equation with time-varying coefficients
IEEE Transactions on Neural Networks
Design and analysis of a general recurrent neural network model for time-varying matrix inversion
IEEE Transactions on Neural Networks
Hi-index | 0.00 |
Our previous work has shown the efficacy and promising performance of continuous-time Zhang dynamics (CTZD) for solving online nonlinear equations, as compared with conventional gradient dynamics (GD). It has also shown that, with linear activation functions used and with step-size being 1, the discrete-time Zhang dynamics (DTZD) reduces to the Newton-Raphson iteration (NRI) (i.e., being a special case of the DTZD model) for static nonlinear equations solving. It is known that the NRI may fail to converge to the theoretical roots of some difficult or special problems. In this paper, the CTZD model and NRI method are investigated comparatively for online solution of static nonlinear equations by performing numerical tests in different situations. Computer testing and simulation results further demonstrate the efficacy and different convergence-performance of the CTZD model (activated by a power-sigmoid function) and NRI for nonlinear equations solving.