Induction of Oblique Decision Trees

This article describes a new system for induction of oblique decision trees. This system, OC1, combines deterministic hill-climbing with two forms of randomization to nd a good oblique split in the form of a hyperplane at each node of a decision tree. Oblique decision tree methodsare tuned especially for domains in which the attributes are numeric, although they can be adapted to symbolic or mixed symbolic numeric attributes. We present extensive empirical studies, using both real and arti cial data, that analyze OC1's ability to construct oblique trees that are smaller and more accurate than their axis-parallel counterparts. We also examine the bene ts of randomization for the construction of oblique decision trees.

• Type: Conference
• Year: 1993
• Where: IJCAI
• Authors: David G. Heath, Simon Kasif, Steven Salzberg
• Keywords: IJCAI 1993, IJCAI 2007, Oblique Decision Trees, Oblique Trees, Symbolic Numeric Attributes