Abstract:In this paper, it is shown that four-layer regular fuzzy neural networks can serve as universal approximators for the sendograph-metric-continuous fuzzy-valued functions. The proof is constructive. We propose a principled method to design four-layer regular fuzzy neural neural network to approximate the target functions. In the previous work, a step function is used as the activation function. To improve the approximation accuracy, in the present work, we also consider using a semi-linear sigmoidal function as the activation function. Then it shows how to design the regular fuzzy neural networks (RFNNs) when the activation functions are the semi-linear sigmoidal function and the step function, respectively. After analyze the approximation accuracy of these two classes of RFNNs, it is found that the former has a much better performance than the latter in approximation accuracy. This conclusion also holds when the target functions satisfy other types of continuity. So the results in this paper can also be used to improve the related work. At last, we give a simulation example to validate the theoretical results.

Approximation capabilities of multilayer fuzzy neural networks on the set of fuzzy-valued functions

Approximation of Level Continuous Fuzzy-Valued Functions by Multilayer Regular Fuzzy Neural Networks

Approximation of fuzzy functions by regular fuzzy neural networks

Approximation of Fuzzy-Valued Functions by Regular Fuzzy Neural Networks and the Accuracy Analysis

Fuzzy RBF Neural Network Based Function Approximation

Function approximation ability of fuzzy neural networks

Approximation Capability of Interpolation Neural Networks

Approximation capability of the convolution methods for fuzzy numbers

Approximation by approximate interpolation neural networks with single hidden layer

Approximation of Fuzzy Numbers Using the Convolution Method

APPROXIMATION MULTIDIMENSION FUCTION WITH FUNCTIONAL NETWORK

Approximation to a Compact Set of Functions by Feedforward Neural Networks.

Exact and Approximate Interpolation for Neural Networks with Single Hidden Layer

Interpolation Mechanism Of Functional Networks

Characterization of Degree of Approximation for Neural Networks with One Hidden Layer

Approximations of Continuous Functionals by Neural Networks with Application to Dynamic Systems

Lp Simultaneous Approximation by Neural Networks with One Hidden Layer

Approximate Interpolation by a Class of Neural Networks in Lebesgue Metric.

Theory and Learning Algorithm of Multi-Dimensional Function Approximation Based on Functional Networks

Approximation and interpolation of deep neural networks

One-variable Interpolation Function Based on Functional Networks