Published:
May 1999
Proceedings:
Proceedings of the Twelfth International Florida Artificial Intelligence Research Society Conference (FLAIRS 1999)
Volume
Issue:
Proceedings of the Twelfth International Florida Artificial Intelligence Research Society Conference (FLAIRS 1999)
Track:
All Papers
Downloads:
Abstract:
We focus on the filter approach of feature selection. We exploit geometrical characterics of the learning set to build an estimation criterion based on a quadratic entropy. The distribution of this criterion is approximately normal, that allows the construction of a non parametrical statistical test to assess the relevance of feature subsets. We use the critical threshold of this test, called the test of Relative Certainty Gain, in a forward selection algorithm. We present some experimental results both on synthetic and natural domains belonging to the UCI database repository, which show significantly improvments on the accuracy estimates.
![](https://aaai.org/wp-content/uploads/2023/06/flairs99.jpg)
FLAIRS
Proceedings of the Twelfth International Florida Artificial Intelligence Research Society Conference (FLAIRS 1999)
ISBN 978-1-57735-080-4
Published by The AAAI Press, Menlo Park, California.