System Modeling with Fuzzy Models: Fundamental Developments and Perspectives
W. Pedrycz
DOI: https://doi.org/10.22111/ijfs.2016.2940
2016-01-01
Iranian Journal of Fuzzy Systems
Abstract:In this study, we offer a general view at the area of fuzzy modeling and fuzzy models, identify the visible development phases and elaborate on a new and promising directions of system modeling by introducing a concept of granular models. Granular models, especially granular fuzzy models constitute an important generalization of existing fuzzy models and, in contrast to the existing models, generate results in the form of information granules (such as intervals, fuzzy sets, rough sets and others). We present a rationale and deliver some key motivating arguments behind the emergence of granular models and discuss their underlying design process. Central to the development of granular models are granular spaces, namely a granular space of parameters of the models and a granular input space. The development of the granular model is completed through an optimal allocation of information granularity, which optimizes criteria of coverage and specificity of granular information. The emergence of granular models of type-2 and type-n, in general, is discussed along with an elaboration on their formation. It is shown that achieving a sound coverage-specificity tradeoff (compromise) is of paramount relevance in the realization of the granular models.