Searching for just a few words should be enough to get started. If you need to make more complex queries, use the tips below to guide you.
Article type: Research Article
Authors: Hashemi, Saeed; *
Affiliations: Faculty of Computer Science, Dalhousie University, 6050 University Ave., Halifax, B3H 1W5, Canada. Tel.: +1 902 494 6441
Correspondence: [*] Postal address: Saeed Hashemi, 1094 Wellington St., #201, Halifax, NS B3H 2Z9, Canada. E-mail: [email protected].
Abstract: Subset selection with a wrapper approach to identify atypical examples can be preferable to a filter approach (which may not be consistent with the classifier in use) but its running time is prohibitive. The fastest available wrappers are quadratic in the number of examples, which is far too expensive for sample subset selection. The presented approach is a linear wrapper method that is roughly 80 times faster than the quadratic wrappers. Atypical points are defined in this paper as the misclassified points that the proposed algorithm (Atypical Sequential Ranking: ASR) finds not useful to the classification task. They may include both outliers and overlapping samples. ASR can identify and rank atypical points in the whole dataset without damaging the prediction accuracy. It is general enough that classifiers without reject option can use it. Experiments on 20 benchmark datasets and 5 classifiers show promising results and confirm that this wrapper method has some advantages and can be used in sample subset selection for atypical detection.
Keywords: atypical data, outlier detection, wrapper method, linear wrapper, sample subset selection
DOI: 10.3233/IDA-2005-9402
Journal: Intelligent Data Analysis, vol. 9, no. 4, pp. 329-345, 2005
IOS Press, Inc.
6751 Tepper Drive
Clifton, VA 20124
USA
Tel: +1 703 830 6300
Fax: +1 703 830 2300
[email protected]
For editorial issues, like the status of your submitted paper or proposals, write to [email protected]
IOS Press
Nieuwe Hemweg 6B
1013 BG Amsterdam
The Netherlands
Tel: +31 20 688 3355
Fax: +31 20 687 0091
[email protected]
For editorial issues, permissions, book requests, submissions and proceedings, contact the Amsterdam office [email protected]
Inspirees International (China Office)
Ciyunsi Beili 207(CapitaLand), Bld 1, 7-901
100025, Beijing
China
Free service line: 400 661 8717
Fax: +86 10 8446 7947
[email protected]
For editorial issues, like the status of your submitted paper or proposals, write to [email protected]
如果您在出版方面需要帮助或有任何建, 件至: [email protected]