Invited Talk World Congress DSA 2011

Learning from Data
An Overview about Decision Tree Induction Methods
with Applications in Marketing, Medicine and Process Control

Michael Kunz
ibai solutions Inc., Fockendorf Germany


Decision Trees are the most popular data mining methods in practice. Decision Trees represent the results of a data mining process in a human understandable manner. The structural hierarchical representation of the outcome allows a human to overlook and understand even a large set of rules. The easy application of decision trees to the data is a further advantage of decision trees over other data mining methods.
There are many decision tree induction methods known from the literature. In the talk we give an overview about the methods. We discuss their potentials and disadvantages. We give an outlook to further developments of decision trees. Finally, we show the advantage of decision trees by several examples from marketing, medicine, and process control.