The past, the present, and the future of fuzzy inductive learning methods

Abstract

The development of computers and network environments has been accompanied with the increase of data stored electrically. The importance of inductive learning methods, which discover knowledge from data, is more increasing. The paper focuses on a representative inductive learning method, 1D3, which acquires a rule set described by a decision tree, and discusses relations between 1D3, other inductive learning methods, and fuzzy inductive learning methods. Lastly, the paper introduces current topics in related study fields and shows future directions of fuzzy inductive learning fields.