Scalable feature mining for sequential data

被引：38

作者：

Lesh, N

Zaki, MJ

Ogihara, M

机构：

[1] Mitsubishi Electr Corp, Res Lab, Cambridge, MA 02139 USA

[2] Rensselaer Polytech Inst, Dept Comp Sci, Troy, NY 12180 USA

[3] Univ Rochester, Dept Comp Sci, Rochester, NY 14627 USA

来源：

IEEE INTELLIGENT SYSTEMS & THEIR APPLICATIONS | 2000年 / 15卷 / 02期

基金：

美国国家科学基金会;

关键词：

D O I：

10.1109/5254.850827

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

To provide good feature selection for sequential domains, FeatureMine was developed. This scalable feature-mining algorithm combines sequence mining and classification algorithms. Tests on three practical domains demonstrate the capability to efficiently handle very large data sets with thousands of items and millions of records.

引用

页码：48 / 56

页数：9

共 12 条

[1]

Caruana R., 1994, MACH LEARN P 1994, P28, DOI 10.1016/B978-1-55860-335-6.50012-X

[2]

Domingos P., 1996, P 13 INT C MACH LEAR, P105

[3]

GOLDING A, 1996, P ML 96 13 INT C MAC, P180

[4]

HART DM, 1992, P 1 INT C AI PLANN S, P106

[5]

Hart P.E., 1973, Pattern recognition and scene analysis

[6]

Kucera H., 1967, COMPUTATIONAL ANAL P

[7]

Kudenko D, 1998, FIFTEENTH NATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE (AAAI-98) AND TENTH CONFERENCE ON INNOVATIVE APPLICATIONS OF ARTIFICAL INTELLIGENCE (IAAI-98) - PROCEEDINGS, P733

[8]

Lee W, 1998, PROCEEDINGS OF THE SEVENTH USENIX SECURITY SYMPOSIUM, P79

[9]

Littlestone N., 1988, Machine Learning, V2, P285, DOI 10.1007/BF00116827

[10]

Liu B, 1998, P 4 INT C KNOWL DISC, P80

← 1 2 →