SE-trees outperform decision trees in noisy domains
Ron Rymon · 1996
As a classifier, a Set Enumeration (SE) tree can be viewed as a generalization of decision trees. It can be shown that, at the cost of a higher complexity, a single SE-tree encapsulates many alternative decision tree structures. An SE-tree enjoys several advantages over decision trees: it allows for domain-based userspecified bias, it supports a flexible tradeoff between the resources allocated to learning and the resulting accuracy, and it can combine knowledge induced from examples with other knowledge sources. In this paper, we empirically demonstrate that SE-trees enjoy a particular advantage over simple decision trees in noisy domains. This advantage manifests itself both in terms of accuracy, and in terms of consistency. Introduction Seminal work by Breiman et al. (1984) and Quinlan (1986), on induction of decision tree classifiers, was followed by a wealth of research by many others to improve upon the basic method. In (Rymon, 1993), we propose a generalization of this framewor...