Skip to Main Content
As they require minimum human intervention, automated modeling approaches find preference in many areas. Fuzzy models - when obtained from system measurements - represent good example of such approaches. Because the emphasis is on its accuracy, the obtained model may exhibit unnecessary complexity which hampers its transparency and computational cost. This paper deals with the issues of transparency and accuracy of data-driven fuzzy models. We present an automated method to simplify data-driven fuzzy models. This method targets simplifying rather than reducing the model. However, model reduction may follow from its simplification if it contains high redundancy.