ABSTRACT
This chapter focuses on the so-called model-driven approach to data analysis. This approach is closely related to the computations performed by neural networks since the latter follow a specific procedure for transforming the original data into a form well suited to solving the original, usually external, problem. The goal of this transformation is to represent the original data using feature vectors that carry sufficient information to solve the external problem. In this chapter, we propose a feature extraction technique based on fuzzy modeling in general and, in particular, using the theory of fuzzy transforms (F-transforms). Generally speaking, features are assumed to be extracted either from raw data or from data models as a subject of training. In both cases, the process of their extraction is associated with a certain way of processing the data. In the case of neural network computing, it is determined by the network architecture; otherwise, it follows from the model chosen to represent the data.
