A New Linguistic Fuzzy PRISM Algorithm

Inductive algorithms for generating fuzzy rules are based on the division of the data set according to fuzzy values created for individual attributes. Its main advantage is the ability to generate interpretable systems. However, it can be degraded if the complexity of the system increase. In this paper, we will propose a new version of the Fuzzy Prism algorithm, which allows for obtaining a set of readable and interpretable linguistic fuzzy rules. It also uses linguistic modifiers, which a llows them to increase their expressiveness and reduce their number.