Recent work has shown that feature subset selection can have a positive affect on the performance of machine learning algorithms. Some algorithms can be slowed or their performance irrelevant or redundant to the learning task. Feature subset selection, then, is a method for enhancing the performance of learning algorithms, reducing the hypothesis search space, and, in some cases, reducing the storage requirement. This paper describes a feature subset selector that uses a correlation based evaluates its effectiveness with three common ML algorithms: a decision tree inducer (C4.5), a naive Bayes classifier, and an instance based learner (IB1). Experiments using a number of standard data sets drawn from real and artificial domains are presented. Feature subset selection gave significant improvement for all three algorithms; C4.5 generated smaller decision trees.
[1]
E. Ghiselli.
Theory of psychological measurement
,
1964
.
[2]
Ron Kohavi,et al.
Wrappers for Feature Subset Selection
,
1997,
Artif. Intell..
[3]
Larry A. Rendell,et al.
A Practical Approach to Feature Selection
,
1992,
ML.
[4]
Thomas G. Dietterich,et al.
Learning with Many Irrelevant Features
,
1991,
AAAI.
[5]
Pat Langley,et al.
Scaling to domains with irrelevant features
,
1997,
COLT 1997.
[6]
Geoffrey Holmes,et al.
Feature selection via the discovery of simple classification rules
,
1995
.