Latent outlier detection and the low precision problem

Fei Wang, Sanjay Chawla, Didi Surian

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

3 Citations (Scopus)

Abstract

The identification of outliers is an intrinsic component of knowledge discovery. However, most outlier detection techniques operate in the observational space, which is often associated with information redundancy and noise. Also, due to the usually high dimensionality of the observational space, the anomalies detected are difficult to comprehend. In this paper we claim that algorithms for discovery of outliers in a latent space will not only lead to more accurate results but potentially provide a natural medium to explain and describe outliers. Specifically, we propose combining Non-Negative Matrix Factorization (NMF) with subspace analysis to discover and interpret outliers. We report on preliminary work towards such an approach.

Original languageEnglish
Title of host publicationProceedings of the ACM SIGKDD Workshop on Outlier Detection and Description, ODD 2013
Pages46-52
Number of pages7
DOIs
Publication statusPublished - 2013
Externally publishedYes
EventACM SIGKDD Workshop on Outlier Detection and Description, ODD 2013 - Chicago, IL
Duration: 11 Aug 201311 Aug 2013

Other

OtherACM SIGKDD Workshop on Outlier Detection and Description, ODD 2013
CityChicago, IL
Period11/8/1311/8/13

Fingerprint

Factorization
Redundancy
Data mining

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Computer Vision and Pattern Recognition

Cite this

Wang, F., Chawla, S., & Surian, D. (2013). Latent outlier detection and the low precision problem. In Proceedings of the ACM SIGKDD Workshop on Outlier Detection and Description, ODD 2013 (pp. 46-52) https://doi.org/10.1145/2500853.2500862

Latent outlier detection and the low precision problem. / Wang, Fei; Chawla, Sanjay; Surian, Didi.

Proceedings of the ACM SIGKDD Workshop on Outlier Detection and Description, ODD 2013. 2013. p. 46-52.

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

Wang, F, Chawla, S & Surian, D 2013, Latent outlier detection and the low precision problem. in Proceedings of the ACM SIGKDD Workshop on Outlier Detection and Description, ODD 2013. pp. 46-52, ACM SIGKDD Workshop on Outlier Detection and Description, ODD 2013, Chicago, IL, 11/8/13. https://doi.org/10.1145/2500853.2500862
Wang F, Chawla S, Surian D. Latent outlier detection and the low precision problem. In Proceedings of the ACM SIGKDD Workshop on Outlier Detection and Description, ODD 2013. 2013. p. 46-52 https://doi.org/10.1145/2500853.2500862
Wang, Fei ; Chawla, Sanjay ; Surian, Didi. / Latent outlier detection and the low precision problem. Proceedings of the ACM SIGKDD Workshop on Outlier Detection and Description, ODD 2013. 2013. pp. 46-52
@inproceedings{f6f64d0721704636a41adcfd18d3b718,
title = "Latent outlier detection and the low precision problem",
abstract = "The identification of outliers is an intrinsic component of knowledge discovery. However, most outlier detection techniques operate in the observational space, which is often associated with information redundancy and noise. Also, due to the usually high dimensionality of the observational space, the anomalies detected are difficult to comprehend. In this paper we claim that algorithms for discovery of outliers in a latent space will not only lead to more accurate results but potentially provide a natural medium to explain and describe outliers. Specifically, we propose combining Non-Negative Matrix Factorization (NMF) with subspace analysis to discover and interpret outliers. We report on preliminary work towards such an approach.",
author = "Fei Wang and Sanjay Chawla and Didi Surian",
year = "2013",
doi = "10.1145/2500853.2500862",
language = "English",
isbn = "9781450323352",
pages = "46--52",
booktitle = "Proceedings of the ACM SIGKDD Workshop on Outlier Detection and Description, ODD 2013",

}

TY - GEN

T1 - Latent outlier detection and the low precision problem

AU - Wang, Fei

AU - Chawla, Sanjay

AU - Surian, Didi

PY - 2013

Y1 - 2013

N2 - The identification of outliers is an intrinsic component of knowledge discovery. However, most outlier detection techniques operate in the observational space, which is often associated with information redundancy and noise. Also, due to the usually high dimensionality of the observational space, the anomalies detected are difficult to comprehend. In this paper we claim that algorithms for discovery of outliers in a latent space will not only lead to more accurate results but potentially provide a natural medium to explain and describe outliers. Specifically, we propose combining Non-Negative Matrix Factorization (NMF) with subspace analysis to discover and interpret outliers. We report on preliminary work towards such an approach.

AB - The identification of outliers is an intrinsic component of knowledge discovery. However, most outlier detection techniques operate in the observational space, which is often associated with information redundancy and noise. Also, due to the usually high dimensionality of the observational space, the anomalies detected are difficult to comprehend. In this paper we claim that algorithms for discovery of outliers in a latent space will not only lead to more accurate results but potentially provide a natural medium to explain and describe outliers. Specifically, we propose combining Non-Negative Matrix Factorization (NMF) with subspace analysis to discover and interpret outliers. We report on preliminary work towards such an approach.

UR - http://www.scopus.com/inward/record.url?scp=84890726684&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84890726684&partnerID=8YFLogxK

U2 - 10.1145/2500853.2500862

DO - 10.1145/2500853.2500862

M3 - Conference contribution

SN - 9781450323352

SP - 46

EP - 52

BT - Proceedings of the ACM SIGKDD Workshop on Outlier Detection and Description, ODD 2013

ER -