Abstract
We have already proposed a fuzzy reasoning method based on the distance between fuzzy sets. However, the distance type fuzzy reasoning method does not have a learning function. This paper discusses the benefits of providing the distance-type fuzzy reasoning method with a learning function. Unlike neural networks, GA, and other conventional approaches, this learning algorithm uses the features of the distance-type fuzzy reasoning method appropriately. Consequently, this algorithm is veg simple and extremely fast, requiring almost no learning time. Finally, the effectiveness of this learning algorithm is verified by simulation studies.
