Image semantic annotation using fuzzy decision trees
Andreea Popescu, Bogdan E. Popescu, Marius Brezovan, Eugen Ganea · 2013
Abstract—One of the methods most commonly used for learning and classification is using decision trees. The greatest advantages that decision trees offer is that, unlike classical trees, they provide a support for handling uncertain data sets. The paper introduces a new algorithm for building fuzzy decision trees and also offers some comparative results, by taking into account other methods. We will present a general overview of the fuzzy decision trees and focus afterwards on the newly introduced algorithm, pointing out that it can be a very useful tool in processing fuzzy data sets by offering good comparative results. I.