Unsupervised learning trajectory anomaly detection algorithm based on deep representation

Without ground-truth data, trajectory anomaly detection is a hard work and the result lacks of interpretability. Moreover, in most current methods, trajectories are represented by geometric features or their low-dimensional linear combination, and some hidden features and high-dimensional combined f...

Full description

Saved in:
Bibliographic Details
Main Authors: Zhongqiu Wang, Guan Yuan, Haoran Pei, Yanmei Zhang, Xiao Liu
Format: Article
Language:English
Published: Wiley 2020-12-01
Series:International Journal of Distributed Sensor Networks
Online Access:https://doi.org/10.1177/1550147720971504
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1832556797453402112
author Zhongqiu Wang
Guan Yuan
Haoran Pei
Yanmei Zhang
Xiao Liu
author_facet Zhongqiu Wang
Guan Yuan
Haoran Pei
Yanmei Zhang
Xiao Liu
author_sort Zhongqiu Wang
collection DOAJ
description Without ground-truth data, trajectory anomaly detection is a hard work and the result lacks of interpretability. Moreover, in most current methods, trajectories are represented by geometric features or their low-dimensional linear combination, and some hidden features and high-dimensional combined features cannot be found efficiently. Meanwhile, traditional methods still cannot get rid of the limitation of space attributes. Therefore, a novel trajectory anomaly detection algorithm is present in this article. Unsupervised learning mechanism is used to overcome nonground-truth problem and deep representation method is used to represent trajectories in a comprehensive way. First, each trajectory is partitioned into segments according to its open angles, then the shallow features at each point of a segment are extracted and. In this way, each segment is represented as a feature sequence. Second, shallow features are integrated into auto-encoder-based deep feature fusion model, and the fusion feature sequences can be extracted. Third, these fused feature sequences are grouped into different clusters using a unsupervised clustering algorithm, and then segments which quite differ from others are detected as anomalies. Finally, comprehensive experiments are conducted on both synthetic and real data sets, which demonstrate the efficiency of our work.
format Article
id doaj-art-1942b400fae3497194d96bd225bfe012
institution Kabale University
issn 1550-1477
language English
publishDate 2020-12-01
publisher Wiley
record_format Article
series International Journal of Distributed Sensor Networks
spelling doaj-art-1942b400fae3497194d96bd225bfe0122025-02-03T05:44:18ZengWileyInternational Journal of Distributed Sensor Networks1550-14772020-12-011610.1177/1550147720971504Unsupervised learning trajectory anomaly detection algorithm based on deep representationZhongqiu Wang0Guan Yuan1Haoran Pei2Yanmei Zhang3Xiao Liu4School of Computer Science and Technology, China University of Mining and Technology, Xuzhou, ChinaJiangsu Key Laboratory of Mine Mechanical and Electrical Equipment, China University of Mining and Technology, Xuzhou, ChinaSchool of Computer Science and Technology, China University of Mining and Technology, Xuzhou, ChinaSchool of Computer Science and Technology, China University of Mining and Technology, Xuzhou, ChinaSchool of Computer Science and Technology, China University of Mining and Technology, Xuzhou, ChinaWithout ground-truth data, trajectory anomaly detection is a hard work and the result lacks of interpretability. Moreover, in most current methods, trajectories are represented by geometric features or their low-dimensional linear combination, and some hidden features and high-dimensional combined features cannot be found efficiently. Meanwhile, traditional methods still cannot get rid of the limitation of space attributes. Therefore, a novel trajectory anomaly detection algorithm is present in this article. Unsupervised learning mechanism is used to overcome nonground-truth problem and deep representation method is used to represent trajectories in a comprehensive way. First, each trajectory is partitioned into segments according to its open angles, then the shallow features at each point of a segment are extracted and. In this way, each segment is represented as a feature sequence. Second, shallow features are integrated into auto-encoder-based deep feature fusion model, and the fusion feature sequences can be extracted. Third, these fused feature sequences are grouped into different clusters using a unsupervised clustering algorithm, and then segments which quite differ from others are detected as anomalies. Finally, comprehensive experiments are conducted on both synthetic and real data sets, which demonstrate the efficiency of our work.https://doi.org/10.1177/1550147720971504
spellingShingle Zhongqiu Wang
Guan Yuan
Haoran Pei
Yanmei Zhang
Xiao Liu
Unsupervised learning trajectory anomaly detection algorithm based on deep representation
International Journal of Distributed Sensor Networks
title Unsupervised learning trajectory anomaly detection algorithm based on deep representation
title_full Unsupervised learning trajectory anomaly detection algorithm based on deep representation
title_fullStr Unsupervised learning trajectory anomaly detection algorithm based on deep representation
title_full_unstemmed Unsupervised learning trajectory anomaly detection algorithm based on deep representation
title_short Unsupervised learning trajectory anomaly detection algorithm based on deep representation
title_sort unsupervised learning trajectory anomaly detection algorithm based on deep representation
url https://doi.org/10.1177/1550147720971504
work_keys_str_mv AT zhongqiuwang unsupervisedlearningtrajectoryanomalydetectionalgorithmbasedondeeprepresentation
AT guanyuan unsupervisedlearningtrajectoryanomalydetectionalgorithmbasedondeeprepresentation
AT haoranpei unsupervisedlearningtrajectoryanomalydetectionalgorithmbasedondeeprepresentation
AT yanmeizhang unsupervisedlearningtrajectoryanomalydetectionalgorithmbasedondeeprepresentation
AT xiaoliu unsupervisedlearningtrajectoryanomalydetectionalgorithmbasedondeeprepresentation