A Neuro-Fuzzy Approach for Data Mining and

Data mining is a technique for discovering hidden knowledge of strategic importance from large databases. This technique is currently being successfully used by a number of real-world applications. Data mining is a tool for increasing the productivity of people trying to build predictive models. Both disciplines have been working on problems of pattern recognition and classification. Rule induction that is a method for deriving a set of rules to classify cases is one of the famous techniques used for data mining. In this paper, we present a new approach to build a fuzzy model from a numerical input–output database through rule induction; using a special type of backpropagation neural network (BPNN). A fuzzy model is derived in three phases: initialization, optimization, and simplification of the fuzzy model. Medical diagnosis of diseases is an important and difficult task in medicine. To verify the effectiveness of the proposed approach, it was applied on three well-known medical diagnosis problems.