Neural Network Operators on fuzzy number valued continuous functions

TL;DR

This paper extends neural network operators to fuzzy number valued continuous functions, analyzing their behavior and establishing Jackson-type approximation results for various types of continuity.

Contribution

It introduces a novel extension of neural network operators to fuzzy-valued functions and provides theoretical approximation results for different continuity types.

Findings

Established Jackson-type approximation results for fuzzy-valued functions.

Analyzed behavior of neural network operators on level, sendograph, and endograph continuous functions.

Extended existing neural network operator theory to fuzzy number contexts.

Abstract

We extend Cardaliaguet-Euvrard neural network operators to the context of fuzzy number valued continuous functions and study their behaviour. We focus on level continuous, sendograph continuous and endograph continuous functions and obtain Jackson-type results in all these cases.