A Building-Block Approach to Hybrid Fuzzy Modeling

In the last years, many authors have solved the data-driven fuzzy modeling (DDFM) problem by combining different techniques into hybrid strategies. Moreover, the hybridization in DDFM has demonstrated to be a successful approach. However, these combinations are made in an ad hoc manner and there exists a lack of mechanisms to make these combinations in an easy and unified manner. This work proposes a new DDFM reference model called METHOD starting from which frameworks for hybridization could be designed. This model offers the functionalities needed to seamlessly combine techniques as building blocks into hybrid strategies for DDFM tasks