Ordinal decision trees based on fuzzy rank entropy

Ordinal classification problems widely exist in the real world, and ordinal decision tree is one of the most important ways of dealing with the ordinal classification problems. In this paper, we introduce the fuzzy rank entropy and design a new ordinal decision tree algorithm (FREMT) based on the fuzzy rank mutual information, which is an extension of ordinal decision tree. The proposed algorithm selects the fuzzy rank mutual information as a split criterion, and it can not only be applied in crisp ordinal set, but also be used in the fuzzy ordinal set The numerical experiments show that its performance is superior to other methods as well as rank mutual information based ordinal decision tree algorithm (REMT).