ACE: Anomaly clustering ensemble for multi-perspective anomaly detection in robot behaviors

Shin Ando, Theerasak Thanomphongphan, Daisuke Hoshino, Yoichi Seki, Einoshin Suzuki

Research output: Chapter in Book/Report/Conference proceedingConference contribution

8 Citations (Scopus)

Abstract

This paper addresses an application of anomaly detection from subsequences of time series (STS) to autonomous robots' behaviors. An important aspect of mining sequential data is selecting the temporal parameters, such as the subsequence length and the degree of smoothing. For example in the task at hand, the patterns of the robot's velocity, which is one of its fundamental features, vary significantly subject to the interval for measuring the displacement. Selecting the time scale and resolution is difficult in unsupervised settings, and is often more critical than the choice of the method. In this paper, we propose an ensemble framework for aggregating anomaly detection from different perspectives, i.e., settings of user-defined, temporal parameters. In the proposed framework, each behavior is labeled whether it is an anomaly in multiple settings. The set of labels are used as meta-features of the respective behaviors. Cluster analysis in a meta-feature space partitions anomalous behaviors pertained to a specific range of parameters. The framework also includes a scalable implementation of the instance-based anomaly detection. We evaluate the proposed framework by ROC analysis, in comparison to conventional ensemble methods for anomaly detection.

Original languageEnglish
Title of host publicationProceedings of the 11th SIAM International Conference on Data Mining, SDM 2011
Pages1-12
Number of pages12
Publication statusPublished - Dec 1 2011
Event11th SIAM International Conference on Data Mining, SDM 2011 - Mesa, AZ, United States
Duration: Apr 28 2011Apr 30 2011

Publication series

NameProceedings of the 11th SIAM International Conference on Data Mining, SDM 2011

Other

Other11th SIAM International Conference on Data Mining, SDM 2011
CountryUnited States
CityMesa, AZ
Period4/28/114/30/11

Fingerprint

Robots
Cluster analysis
Labels
Time series

All Science Journal Classification (ASJC) codes

  • Software

Cite this

Ando, S., Thanomphongphan, T., Hoshino, D., Seki, Y., & Suzuki, E. (2011). ACE: Anomaly clustering ensemble for multi-perspective anomaly detection in robot behaviors. In Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011 (pp. 1-12). (Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011).

ACE : Anomaly clustering ensemble for multi-perspective anomaly detection in robot behaviors. / Ando, Shin; Thanomphongphan, Theerasak; Hoshino, Daisuke; Seki, Yoichi; Suzuki, Einoshin.

Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011. 2011. p. 1-12 (Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Ando, S, Thanomphongphan, T, Hoshino, D, Seki, Y & Suzuki, E 2011, ACE: Anomaly clustering ensemble for multi-perspective anomaly detection in robot behaviors. in Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011. Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011, pp. 1-12, 11th SIAM International Conference on Data Mining, SDM 2011, Mesa, AZ, United States, 4/28/11.
Ando S, Thanomphongphan T, Hoshino D, Seki Y, Suzuki E. ACE: Anomaly clustering ensemble for multi-perspective anomaly detection in robot behaviors. In Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011. 2011. p. 1-12. (Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011).
Ando, Shin ; Thanomphongphan, Theerasak ; Hoshino, Daisuke ; Seki, Yoichi ; Suzuki, Einoshin. / ACE : Anomaly clustering ensemble for multi-perspective anomaly detection in robot behaviors. Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011. 2011. pp. 1-12 (Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011).
@inproceedings{f336fb82d69041a988affe067726bb4c,
title = "ACE: Anomaly clustering ensemble for multi-perspective anomaly detection in robot behaviors",
abstract = "This paper addresses an application of anomaly detection from subsequences of time series (STS) to autonomous robots' behaviors. An important aspect of mining sequential data is selecting the temporal parameters, such as the subsequence length and the degree of smoothing. For example in the task at hand, the patterns of the robot's velocity, which is one of its fundamental features, vary significantly subject to the interval for measuring the displacement. Selecting the time scale and resolution is difficult in unsupervised settings, and is often more critical than the choice of the method. In this paper, we propose an ensemble framework for aggregating anomaly detection from different perspectives, i.e., settings of user-defined, temporal parameters. In the proposed framework, each behavior is labeled whether it is an anomaly in multiple settings. The set of labels are used as meta-features of the respective behaviors. Cluster analysis in a meta-feature space partitions anomalous behaviors pertained to a specific range of parameters. The framework also includes a scalable implementation of the instance-based anomaly detection. We evaluate the proposed framework by ROC analysis, in comparison to conventional ensemble methods for anomaly detection.",
author = "Shin Ando and Theerasak Thanomphongphan and Daisuke Hoshino and Yoichi Seki and Einoshin Suzuki",
year = "2011",
month = "12",
day = "1",
language = "English",
isbn = "9780898719925",
series = "Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011",
pages = "1--12",
booktitle = "Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011",

}

TY - GEN

T1 - ACE

T2 - Anomaly clustering ensemble for multi-perspective anomaly detection in robot behaviors

AU - Ando, Shin

AU - Thanomphongphan, Theerasak

AU - Hoshino, Daisuke

AU - Seki, Yoichi

AU - Suzuki, Einoshin

PY - 2011/12/1

Y1 - 2011/12/1

N2 - This paper addresses an application of anomaly detection from subsequences of time series (STS) to autonomous robots' behaviors. An important aspect of mining sequential data is selecting the temporal parameters, such as the subsequence length and the degree of smoothing. For example in the task at hand, the patterns of the robot's velocity, which is one of its fundamental features, vary significantly subject to the interval for measuring the displacement. Selecting the time scale and resolution is difficult in unsupervised settings, and is often more critical than the choice of the method. In this paper, we propose an ensemble framework for aggregating anomaly detection from different perspectives, i.e., settings of user-defined, temporal parameters. In the proposed framework, each behavior is labeled whether it is an anomaly in multiple settings. The set of labels are used as meta-features of the respective behaviors. Cluster analysis in a meta-feature space partitions anomalous behaviors pertained to a specific range of parameters. The framework also includes a scalable implementation of the instance-based anomaly detection. We evaluate the proposed framework by ROC analysis, in comparison to conventional ensemble methods for anomaly detection.

AB - This paper addresses an application of anomaly detection from subsequences of time series (STS) to autonomous robots' behaviors. An important aspect of mining sequential data is selecting the temporal parameters, such as the subsequence length and the degree of smoothing. For example in the task at hand, the patterns of the robot's velocity, which is one of its fundamental features, vary significantly subject to the interval for measuring the displacement. Selecting the time scale and resolution is difficult in unsupervised settings, and is often more critical than the choice of the method. In this paper, we propose an ensemble framework for aggregating anomaly detection from different perspectives, i.e., settings of user-defined, temporal parameters. In the proposed framework, each behavior is labeled whether it is an anomaly in multiple settings. The set of labels are used as meta-features of the respective behaviors. Cluster analysis in a meta-feature space partitions anomalous behaviors pertained to a specific range of parameters. The framework also includes a scalable implementation of the instance-based anomaly detection. We evaluate the proposed framework by ROC analysis, in comparison to conventional ensemble methods for anomaly detection.

UR - http://www.scopus.com/inward/record.url?scp=84857164317&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84857164317&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84857164317

SN - 9780898719925

T3 - Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011

SP - 1

EP - 12

BT - Proceedings of the 11th SIAM International Conference on Data Mining, SDM 2011

ER -