Adaptive sampling and learning for unsupervised outlier detection

Zhiruo Zhao; Chilukuri K. Mohan; Kishan G. Mehrotra

Adaptive sampling and learning for unsupervised outlier detection

Zhiruo Zhao, Chilukuri K. Mohan, Kishan G. Mehrotra

Department of Electrical Engineering & Computer Science

Research output: Chapter in Book/Entry/Poem › Conference contribution

3 Scopus citations

Abstract

Unsupervised outlier detection algorithms often suffer from high false positive detection rates. Ensemble approaches can be used to address this problem. This paper proposes a novel ensemble method which adopts the use of an adaptive sampling approach, and combines the outputs of individual anomaly detection algorithms by a weighted majority voting rule in a complete unsupervised context. Simulations on well-known benchmark problems show substantial improvement in performance.

Original language	English (US)
Title of host publication	Proceedings of the 29th International Florida Artificial Intelligence Research Society Conference, FLAIRS 2016
Publisher	AAAI Press
Pages	460-465
Number of pages	6
ISBN (Electronic)	9781577357568
State	Published - 2016
Event	29th International Florida Artificial Intelligence Research Society Conference, FLAIRS 2016 - Key Largo, United States Duration: May 16 2016 → May 18 2016

Other

Other	29th International Florida Artificial Intelligence Research Society Conference, FLAIRS 2016
Country/Territory	United States
City	Key Largo
Period	5/16/16 → 5/18/16

ASJC Scopus subject areas

Software
Artificial Intelligence
Computer Networks and Communications

Cite this

Zhao, Z, Mohan, CK & Mehrotra, KG 2016, Adaptive sampling and learning for unsupervised outlier detection. in Proceedings of the 29th International Florida Artificial Intelligence Research Society Conference, FLAIRS 2016. AAAI Press, pp. 460-465, 29th International Florida Artificial Intelligence Research Society Conference, FLAIRS 2016, Key Largo, United States, 5/16/16.

@inproceedings{49476e758de345e88a82ba3531944060,

title = "Adaptive sampling and learning for unsupervised outlier detection",

abstract = "Unsupervised outlier detection algorithms often suffer from high false positive detection rates. Ensemble approaches can be used to address this problem. This paper proposes a novel ensemble method which adopts the use of an adaptive sampling approach, and combines the outputs of individual anomaly detection algorithms by a weighted majority voting rule in a complete unsupervised context. Simulations on well-known benchmark problems show substantial improvement in performance.",

author = "Zhiruo Zhao and Mohan, {Chilukuri K.} and Mehrotra, {Kishan G.}",

year = "2016",

language = "English (US)",

pages = "460--465",

booktitle = "Proceedings of the 29th International Florida Artificial Intelligence Research Society Conference, FLAIRS 2016",

publisher = "AAAI Press",

note = "29th International Florida Artificial Intelligence Research Society Conference, FLAIRS 2016 ; Conference date: 16-05-2016 Through 18-05-2016",

}

TY - GEN

T1 - Adaptive sampling and learning for unsupervised outlier detection

AU - Zhao, Zhiruo

AU - Mohan, Chilukuri K.

AU - Mehrotra, Kishan G.

PY - 2016

Y1 - 2016

N2 - Unsupervised outlier detection algorithms often suffer from high false positive detection rates. Ensemble approaches can be used to address this problem. This paper proposes a novel ensemble method which adopts the use of an adaptive sampling approach, and combines the outputs of individual anomaly detection algorithms by a weighted majority voting rule in a complete unsupervised context. Simulations on well-known benchmark problems show substantial improvement in performance.

AB - Unsupervised outlier detection algorithms often suffer from high false positive detection rates. Ensemble approaches can be used to address this problem. This paper proposes a novel ensemble method which adopts the use of an adaptive sampling approach, and combines the outputs of individual anomaly detection algorithms by a weighted majority voting rule in a complete unsupervised context. Simulations on well-known benchmark problems show substantial improvement in performance.

UR - http://www.scopus.com/inward/record.url?scp=85003828595&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85003828595&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85003828595

SP - 460

EP - 465

BT - Proceedings of the 29th International Florida Artificial Intelligence Research Society Conference, FLAIRS 2016

PB - AAAI Press

T2 - 29th International Florida Artificial Intelligence Research Society Conference, FLAIRS 2016

Y2 - 16 May 2016 through 18 May 2016

ER -

Adaptive sampling and learning for unsupervised outlier detection

Abstract

Other

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this