A Comprehensible Approach to Develop Fuzzy Decision Trees

Abstract

With the immense increase of the data in various fields, interpreting the data into useful information has become a tedious job. Design of models to handle the problem is essential. This paper discusses the methods that handle uncertain information with continuous data and deliver comprehensible classification model. We investigate fuzzy decision tree as a method for classification problems and axiomatic fuzzy set for building fuzzy sets (membership functions) . To select the best available test attributes of fuzzy decision trees we use a generalized Shannon Entropy. The problems connected with this generalization arised from fuzzy domain are discussed and some alternatives are proposed.