Axiomatic approach to feature subset selection based on relevance

Author

Wang, Hui ; Bell, David ; Murtagh, Fionn

Author_Institution

Fac. of Inf., Ulster Univ., Newtownabbey, UK

Volume

21

Issue

3

fYear

1999

fDate

3/1/1999 12:00:00 AM

Firstpage

271

Lastpage

277

Abstract

Relevance has traditionally been linked with feature subset selection, but formalization of this link has not been attempted. In this paper, we propose two axioms for feature subset selection-sufficiency axiom and necessity axiom-based on which this link is formalized: The expected feature subset is the one which maximizes relevance. Finding the expected feature subset turns out to be NP-hard. We then devise a heuristic algorithm to find the expected subset which has a polynomial time complexity. The experimental results show that the algorithm finds good enough subset of features which, when presented to C4.5, results in better prediction accuracy

Keywords

computational complexity; feature extraction; heuristic programming; optimisation; C4.5; NP-hard problem; expected feature subset; feature subset selection; heuristic algorithm; polynomial time complexity; relevance; relevance maximization; Accuracy; Entropy; Filters; Frequency selective surfaces; Heuristic algorithms; Machine learning; Machine learning algorithms; Polynomials; Solids; Statistics;

fLanguage

English

Journal_Title

Pattern Analysis and Machine Intelligence, IEEE Transactions on

Publisher

ieee

ISSN

0162-8828

Type

jour

DOI

10.1109/34.754624

Filename

754624