LEARNING FUZZY DECISION TREES FROM SEQUENTIAL AND INCOMPLETE DATA

摘要

In this paper we present a way of constructing and using fuzzy decision trees for machine learning from sequential and incomplete data. We develop a theory and technique for processing such data. At first, we introduce a decision tree with fuzzy attributes and class the properties of fuzzy attributes and classes, where we will introduce the fuzzy division of objects in classes in a different way than that used in the literature. Then we will also introduce properties of fuzzy attribute value to enrich the fuzziness of our decision trees.