Ordinal Decision Trees Based On Fuzzy Rank Entropy And Comparative Study

Ordinal classification tasks widely exist in real word life. Rank entropy based ordinaldecision tree is one of the most important ways of dealing with ordinal classification problems.Based on this work, a fuzzy ordinal decision tree algorithm is proposed in this paper byintroducing fuzzy rank entropy, which employs fuzzy rank mutual information to select theexpanded attributes, the proposed alogithm is an extension of ordinal decision tree algorithm.It can efficiently deal with the questions with preference ordered relation in our life.Both of them have their advantages and disadvantages. In this paper we apply C4.5、RT、REMT and FREMT on15datasets and analyze four different decision trees from threeaspects: the complexity of the tree, the testing accuracy and the selection of expandedattributes. The comparative study provides a meaningful exploration for selecting differentalgorithms to solve different problems.