Robustness Analyses and Optimal Sampling Gap of Recurrent Neural Network for Dynamic Matrix Pseudoinversion

This study analyses the robustness and convergence characteristics of a neural network. First, a special class of recurrent neural network (RNN), termed a continuous-time Zhang neural network (CTZNN) model, is presented and investigated for dynamic matrix pseudoinversion. Theoretical analysis of the CTZNN model demonstrates that it has good robustness against various types of noise. In addition, considering the requirements of digital implementation and online computation, the optimal sampling gap for a discrete-time Zhang neural network (DTZNN) model under noisy environments is proposed. Finally, experimental results are presented, which further substantiate the theoretical analyses and demonstrate the effectiveness of the proposed ZNN models for computing a dynamic matrix pseudoinverse under noisy environments.

THE EFFECTIVENESS OF THE LEARNING ALGORITHM OF RADIAL BASIS NETWORKS WITH RELATION TO THE TRANSFER FUNCTIONS APPLIED ON THE EXAMPLE OF MAPPING OF THE LIE LAND OF ZIELONA GORA CITY

The article presents problems connected to the use of radial basis networks for the approximation of the ground surface. The main goal of this paper is to research into the precision of topographic profile representation with relation to the transfer functions applied. The paper contains a description of the structure of a radial basis network and a description of networks learning by means of the hybrid method with the use of the notion of the Green matrix pseudoinverse. Special attention was given to the problem of a choice of transfer functions: the Gauss function, the exponential function, the Hardy function, the spliced function of the third and fourth degree as well as bicentral functions with an independent slope and rotation. the result of this article is an example of the operation of a network with relation the transfer functions under discussion.