Neuro–Fuzzy Systems
Detlef D. Nauck, Rudolf Kruse · 2020
Neuro–fuzzy systems are a very important and popular approach in the area of hybrid systems. In this chapter we discuss features of neuro–fuzzy models and how they can be used to create fuzzy models from data. There are several ways to learn fuzzy rules and fuzzy sets by means of learning rules derived from neural network theory. We present several techniques to train a fuzzy model and consider the application of neuro–fuzzy methods in the areas of function approximation, control, and classification. We look at some special neuro–fuzzy approaches and discuss their features and potentials. In addition to neuro–fuzzy methods that are used to support the design of fuzzy models, we also take a look at fuzzy neural networks. They are fuzzifications of standard neural networks that can be used to implement mappings from fuzzy numbers to fuzzy numbers.