Fuzzy decision trees

Decision trees are arguably one of the most popular choices for learning and reasoning systems, especially when it comes to learning from discrete valued (feature based) examples. Because of the success they acquired in their area, there have been many attempts to generalize the method to better suit working with real-valued, numerical attributes, but also for missing values or even numerical outcomes (i.e. for regression tasks). In this paper, we will present a few methods aiming to combine the increased readability of decision trees with the ability to deal with numeric or inaccurate data provided by fuzzy reasoning.