Are We Losing Accuracy While Gaining Confidence in Induced Rules -
An Assessment of PrIL

F. Özden Gür Ali, GE Corporate Research and Development and William A. Wallace, Rensselaer Polytechnic Institute

Probabilistic Inductive Learning (PrIL), is a tree induction algorithm that provides a minimum correct classification level with a specified confidence for each rule in the decision tree. This feature is particularly useful in uncertain environments where the decisions are based on the induced rules. This paper provides a concise description of (the extended) PrIL, and demonstrates that its performance is as good as best results in the machine learning literature, using datasets from the data repository at UC Irvine.

This page is copyrighted by AAAI. All rights reserved. Your use of this site constitutes acceptance of all of AAAI's terms and conditions and privacy policy.