Evaluation of Attribute Selection Measures in Decision Tree induction
Tu-Bao Ho, TrongDung Nguyen · Industrial and Engineering Applications of Artificial Intelligence and Expert Systems · 2022
Many attribute selection measures have been proposed for decision tree induction, but little was known regarding their experimental comparative evaluation. This paper aimed at two following objectives. One was the introduction of a new attribute selection measure ( R -measure) for decision tree induction. The second was an experimental evaluation of the performance of inductive systems using R -measure and eight different attribute selection measures used by machine learning community: gain-ratio, gini-index, gini’-index, Relief, J-measure, d N distance, relevance, χ 2 .