Fuzzy Clustering-based Anomaly Detection for Distributed Multi-view Data

Hongwei Wang, Tianhua Chen, Hongtao Wang, Xuqiang Shao, Pan Su

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

Abstract

Anomaly detection aims to identify the abnormal instances, whose behavior deviates significantly from the others. Nowadays owing to the existence of diverse data generation sources, different attributes of the same instances may be located on distributed parties forming a multi-view dataset. Thus multiview anomaly detection has become a key task to discover outliers across various views. Traditionally, to perform multiview anomaly detection, one needs to centralize data instances from all views into a single machine. However, in many real-world scenarios, it is impractical to send data from diverse views to a master machine due to the privacy issues. Inspired by this, we propose a fuzzy clustering based distributed approach for multiview anomaly detection that simultaneously learns a membership degree matrix for each view and then detects anomalies for all parties. Specifically, we first introduce a combined fuzzy cmeans clustering method for multi-view data and then design an anomaly measurement criterion to quantify the abnormal score from membership degree matrix. To solve the proposed model, a protocol is provided to unify all parties performing a well-designed optimization in an iterative way. Experiments on three datasets with different anomaly settings demonstrate the effectiveness of our approach.
Original languageEnglish
Title of host publication2018 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE 2018)
PublisherIEEE
Pages1-7
Number of pages7
ISBN (Electronic)9781509060207
ISBN (Print)9781509060214
DOIs
Publication statusPublished - 15 Oct 2018
EventIEEE International Conference on Fuzzy Systems - Rio de Janeiro, Brazil
Duration: 8 Jul 201813 Jul 2018
http://www.ecomp.poli.br/~wcci2018/ (Link to Conference Website)

Conference

ConferenceIEEE International Conference on Fuzzy Systems
Abbreviated titleFUZZ-IEEE 2018
CountryBrazil
CityRio de Janeiro
Period8/07/1813/07/18
Internet address

Fingerprint

Fuzzy clustering
Experiments

Cite this

Wang, H., Chen, T., Wang, H., Shao, X., & Su, P. (2018). Fuzzy Clustering-based Anomaly Detection for Distributed Multi-view Data. In 2018 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE 2018) (pp. 1-7). [8491464] IEEE. https://doi.org/10.1109/FUZZ-IEEE.2018.8491464
Wang, Hongwei ; Chen, Tianhua ; Wang, Hongtao ; Shao, Xuqiang ; Su, Pan . / Fuzzy Clustering-based Anomaly Detection for Distributed Multi-view Data. 2018 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE 2018). IEEE, 2018. pp. 1-7
@inproceedings{89821f3caf494fb7be4a570cc41deab1,
title = "Fuzzy Clustering-based Anomaly Detection for Distributed Multi-view Data",
abstract = "Anomaly detection aims to identify the abnormal instances, whose behavior deviates significantly from the others. Nowadays owing to the existence of diverse data generation sources, different attributes of the same instances may be located on distributed parties forming a multi-view dataset. Thus multiview anomaly detection has become a key task to discover outliers across various views. Traditionally, to perform multiview anomaly detection, one needs to centralize data instances from all views into a single machine. However, in many real-world scenarios, it is impractical to send data from diverse views to a master machine due to the privacy issues. Inspired by this, we propose a fuzzy clustering based distributed approach for multiview anomaly detection that simultaneously learns a membership degree matrix for each view and then detects anomalies for all parties. Specifically, we first introduce a combined fuzzy cmeans clustering method for multi-view data and then design an anomaly measurement criterion to quantify the abnormal score from membership degree matrix. To solve the proposed model, a protocol is provided to unify all parties performing a well-designed optimization in an iterative way. Experiments on three datasets with different anomaly settings demonstrate the effectiveness of our approach.",
keywords = "Anomaly Detection, distributed databases, Optimization, protocols, data privacy, privacy, Entropy, fuzzy set theory, learning (artificial intelligence), matrix algebra, pattern clustering",
author = "Hongwei Wang and Tianhua Chen and Hongtao Wang and Xuqiang Shao and Pan Su",
year = "2018",
month = "10",
day = "15",
doi = "10.1109/FUZZ-IEEE.2018.8491464",
language = "English",
isbn = "9781509060214",
pages = "1--7",
booktitle = "2018 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE 2018)",
publisher = "IEEE",

}

Wang, H, Chen, T, Wang, H, Shao, X & Su, P 2018, Fuzzy Clustering-based Anomaly Detection for Distributed Multi-view Data. in 2018 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE 2018)., 8491464, IEEE, pp. 1-7, IEEE International Conference on Fuzzy Systems, Rio de Janeiro, Brazil, 8/07/18. https://doi.org/10.1109/FUZZ-IEEE.2018.8491464

Fuzzy Clustering-based Anomaly Detection for Distributed Multi-view Data. / Wang, Hongwei; Chen, Tianhua; Wang, Hongtao; Shao, Xuqiang ; Su, Pan .

2018 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE 2018). IEEE, 2018. p. 1-7 8491464.

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

TY - GEN

T1 - Fuzzy Clustering-based Anomaly Detection for Distributed Multi-view Data

AU - Wang, Hongwei

AU - Chen, Tianhua

AU - Wang, Hongtao

AU - Shao, Xuqiang

AU - Su, Pan

PY - 2018/10/15

Y1 - 2018/10/15

N2 - Anomaly detection aims to identify the abnormal instances, whose behavior deviates significantly from the others. Nowadays owing to the existence of diverse data generation sources, different attributes of the same instances may be located on distributed parties forming a multi-view dataset. Thus multiview anomaly detection has become a key task to discover outliers across various views. Traditionally, to perform multiview anomaly detection, one needs to centralize data instances from all views into a single machine. However, in many real-world scenarios, it is impractical to send data from diverse views to a master machine due to the privacy issues. Inspired by this, we propose a fuzzy clustering based distributed approach for multiview anomaly detection that simultaneously learns a membership degree matrix for each view and then detects anomalies for all parties. Specifically, we first introduce a combined fuzzy cmeans clustering method for multi-view data and then design an anomaly measurement criterion to quantify the abnormal score from membership degree matrix. To solve the proposed model, a protocol is provided to unify all parties performing a well-designed optimization in an iterative way. Experiments on three datasets with different anomaly settings demonstrate the effectiveness of our approach.

AB - Anomaly detection aims to identify the abnormal instances, whose behavior deviates significantly from the others. Nowadays owing to the existence of diverse data generation sources, different attributes of the same instances may be located on distributed parties forming a multi-view dataset. Thus multiview anomaly detection has become a key task to discover outliers across various views. Traditionally, to perform multiview anomaly detection, one needs to centralize data instances from all views into a single machine. However, in many real-world scenarios, it is impractical to send data from diverse views to a master machine due to the privacy issues. Inspired by this, we propose a fuzzy clustering based distributed approach for multiview anomaly detection that simultaneously learns a membership degree matrix for each view and then detects anomalies for all parties. Specifically, we first introduce a combined fuzzy cmeans clustering method for multi-view data and then design an anomaly measurement criterion to quantify the abnormal score from membership degree matrix. To solve the proposed model, a protocol is provided to unify all parties performing a well-designed optimization in an iterative way. Experiments on three datasets with different anomaly settings demonstrate the effectiveness of our approach.

KW - Anomaly Detection

KW - distributed databases

KW - Optimization

KW - protocols

KW - data privacy

KW - privacy

KW - Entropy

KW - fuzzy set theory

KW - learning (artificial intelligence)

KW - matrix algebra

KW - pattern clustering

U2 - 10.1109/FUZZ-IEEE.2018.8491464

DO - 10.1109/FUZZ-IEEE.2018.8491464

M3 - Conference contribution

SN - 9781509060214

SP - 1

EP - 7

BT - 2018 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE 2018)

PB - IEEE

ER -

Wang H, Chen T, Wang H, Shao X, Su P. Fuzzy Clustering-based Anomaly Detection for Distributed Multi-view Data. In 2018 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE 2018). IEEE. 2018. p. 1-7. 8491464 https://doi.org/10.1109/FUZZ-IEEE.2018.8491464