Work

A System for Induction of Oblique Decision Trees

Year: 1994

Type: article

Abstract: This article describes a new system for induction ofoblique decision trees. This system, OC1, combines deterministic hill-climbing with two forms of randomization to find a goodoblique split (in the form of a hyperplane) at each node of a decisiontree. Oblique decision tree methods are tuned especially for domains in which the attributes are numeric, although they can be adapted to symbolic or mixed symbolic/numeric attributes. We presentextensive empirical studies, using both real and artificial data, thatanalyze OC1's ability to construct oblique trees that are smaller and more accurate than their axis-parallel counterparts. We also examinethe benefits of randomization for the construction of oblique decisiontrees.

Source: Journal of Artificial Intelligence Research

DOI: 10.1613/jair.63

Authors: S. K. Murthy, S. Kasif, S. Salzberg

Institution: Johns Hopkins University

Language: English

Topic: Data Mining Algorithms and Applications

Subfield: Information Systems

Field: Computer Science

Domain: Physical Sciences

Open Access status: diamond

Funders: National Science Foundation, Yale University, Ames Research Center

Awards: Constraint Solving and Matching: Parallel Algorithms and Applications, A Geometric Framework for the Exploration and Analysis of Astrophysical Data, Knowledge Discovery for Molecular Biology: Algorithms and Experiments