In this paper, we are going to propose an approach to design linguistic models which are accurate to a high degree and may be suitably interpreted. This approach will be based on the development of a Hierarchical System of Linguistic Rules learning methodology. This methodology has been thought as a refinement of simple linguistic models which, preserving their descriptive power, introduces small changes to increase their accuracy. To do so, we extend the structure of the Knowledge Base of Fuzzy Rule Base Systems in a hierarchical way, in order to make it more flexible. This flexibilization will allow us to have linguistic rules defined over linguistic partitions with different granularity levels, and thus to improve the modeling of those problem subspaces where the former models have bad performance.
- Genetic algorithms
- Hierarchical knowledge base
- Hierarchical linguistic partitions
- Linguistic modeling
- Mamdani-type fuzzy rule-based systems
- Rule selection