𝔖 Bobbio Scriptorium
✦   LIBER   ✦

Can neural nets be universal approximators for fuzzy functions?

✍ Scribed by J.J. Buckley; Yoichi Hayashi

Publisher
Elsevier Science
Year
1999
Tongue
English
Weight
565 KB
Volume
101
Category
Article
ISSN
0165-0114

No coin nor oath required. For personal study only.

✦ Synopsis

We first argue that the extension principle is too computationally involved to be an efficient way for a computer to evaluate fuzzy functions. We then suggest using c~-cuts and interval arithmetic to compute the values of fuzzy functions. Using this method of computing fuzzy functions, we then show that neural nets are universal approximators for (computable) fuzzy functions, when we only input non-negative, or non-positive, fuzzy numbers.

πŸ“œ SIMILAR VOLUMES

Universal approximators for fuzzy functi
Universal approximators for fuzzy functions
✍ James J. Buckley; Thomas Feuring πŸ“‚ Article πŸ“… 2000 πŸ› Elsevier Science 🌐 English βš– 461 KB

We show how to construct a large class of universal approximators for fuzzy functions (which continuously map fuzzy numbers into fuzzy numbers and are the extension principle extensions of continuous real-valued functions). One important application is that layered, feedforward, neural nets, with re

Fuzzy systems are universal approximator
Fuzzy systems are universal approximators for a smooth function and its derivatives
✍ Vladik Kreinovich; Hung T. Nguyen; Yeung Yam πŸ“‚ Article πŸ“… 2000 πŸ› John Wiley and Sons 🌐 English βš– 96 KB

One of the reasons why fuzzy methodology is successful is that fuzzy systems are universal approximators, i.e., we can approximate an arbitrary continuous function within Ε½ . any given accuracy by a fuzzy system. In some practical applications e.g., in control , it is Ε½ desirable to approximate not