Deep weakly-supervised anomaly detection

Recent semi-supervised anomaly detection methods that are trained using small labeled anomaly examples and large unlabeled data (mostly normal data) have shown largely improved performance over unsupervised methods. However, these methods often focus on fitting abnormalities illustrated by the given...

Full description

Saved in:
Bibliographic Details
Main Authors: PANG, Guansong, SHEN, Chunhua, JIN, Huidong, VAN DEN HENGEL, Anton
Format: text
Language:English
Published: Institutional Knowledge at Singapore Management University 2023
Subjects:
Online Access:https://ink.library.smu.edu.sg/sis_research/8411
https://ink.library.smu.edu.sg/context/sis_research/article/9414/viewcontent/Deep_Weakly_supervised_Anomaly_Detection.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Singapore Management University
Language: English
id sg-smu-ink.sis_research-9414
record_format dspace
spelling sg-smu-ink.sis_research-94142024-01-09T03:47:22Z Deep weakly-supervised anomaly detection PANG, Guansong SHEN, Chunhua JIN, Huidong VAN DEN HENGEL, Anton Recent semi-supervised anomaly detection methods that are trained using small labeled anomaly examples and large unlabeled data (mostly normal data) have shown largely improved performance over unsupervised methods. However, these methods often focus on fitting abnormalities illustrated by the given anomaly examples only (i.e., seen anomalies), and consequently they fail to generalize to those that are not, i.e., new types/classes of anomaly unseen during training. To detect both seen and unseen anomalies, we introduce a novel deep weakly-supervised approach, namely Pairwise Relation prediction Network (PReNet), that learns pairwise relation features and anomaly scores by predicting the relation of any two randomly sampled training instances, in which the pairwise relation can be anomaly-anomaly, anomaly-unlabeled, or unlabeled-unlabeled. Since unlabeled instances are mostly normal, the relation prediction enforces a joint learning of anomaly-anomaly, anomaly-normal, and normal-normal pairwise discriminative patterns, respectively. PReNet can then detect any seen/unseen abnormalities that fit the learned pairwise abnormal patterns, or deviate from the normal patterns. Further, this pairwise approach also seamlessly and significantly augments the training anomaly data. Empirical results on 12 real-world datasets show that PReNet significantly outperforms nine competing methods in detecting seen and unseen anomalies. We also theoretically and empirically justify the robustness of our model w.r.t. anomaly contamination in the unlabeled data. The code is available at https://github.com/mala-lab/PReNet. 2023-08-01T07:00:00Z text application/pdf https://ink.library.smu.edu.sg/sis_research/8411 info:doi/10.1145/3580305.3599302 https://ink.library.smu.edu.sg/context/sis_research/article/9414/viewcontent/Deep_Weakly_supervised_Anomaly_Detection.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ Research Collection School Of Computing and Information Systems eng Institutional Knowledge at Singapore Management University Anomaly detection Anomaly detection methods Deep learning; Intrusion-Detection Learn+ Performance Semi-supervised Type class Unlabeled data Unsupervised method Databases and Information Systems
institution Singapore Management University
building SMU Libraries
continent Asia
country Singapore
Singapore
content_provider SMU Libraries
collection InK@SMU
language English
topic Anomaly detection
Anomaly detection methods
Deep learning; Intrusion-Detection
Learn+
Performance
Semi-supervised
Type class
Unlabeled data
Unsupervised method
Databases and Information Systems
spellingShingle Anomaly detection
Anomaly detection methods
Deep learning; Intrusion-Detection
Learn+
Performance
Semi-supervised
Type class
Unlabeled data
Unsupervised method
Databases and Information Systems
PANG, Guansong
SHEN, Chunhua
JIN, Huidong
VAN DEN HENGEL, Anton
Deep weakly-supervised anomaly detection
description Recent semi-supervised anomaly detection methods that are trained using small labeled anomaly examples and large unlabeled data (mostly normal data) have shown largely improved performance over unsupervised methods. However, these methods often focus on fitting abnormalities illustrated by the given anomaly examples only (i.e., seen anomalies), and consequently they fail to generalize to those that are not, i.e., new types/classes of anomaly unseen during training. To detect both seen and unseen anomalies, we introduce a novel deep weakly-supervised approach, namely Pairwise Relation prediction Network (PReNet), that learns pairwise relation features and anomaly scores by predicting the relation of any two randomly sampled training instances, in which the pairwise relation can be anomaly-anomaly, anomaly-unlabeled, or unlabeled-unlabeled. Since unlabeled instances are mostly normal, the relation prediction enforces a joint learning of anomaly-anomaly, anomaly-normal, and normal-normal pairwise discriminative patterns, respectively. PReNet can then detect any seen/unseen abnormalities that fit the learned pairwise abnormal patterns, or deviate from the normal patterns. Further, this pairwise approach also seamlessly and significantly augments the training anomaly data. Empirical results on 12 real-world datasets show that PReNet significantly outperforms nine competing methods in detecting seen and unseen anomalies. We also theoretically and empirically justify the robustness of our model w.r.t. anomaly contamination in the unlabeled data. The code is available at https://github.com/mala-lab/PReNet.
format text
author PANG, Guansong
SHEN, Chunhua
JIN, Huidong
VAN DEN HENGEL, Anton
author_facet PANG, Guansong
SHEN, Chunhua
JIN, Huidong
VAN DEN HENGEL, Anton
author_sort PANG, Guansong
title Deep weakly-supervised anomaly detection
title_short Deep weakly-supervised anomaly detection
title_full Deep weakly-supervised anomaly detection
title_fullStr Deep weakly-supervised anomaly detection
title_full_unstemmed Deep weakly-supervised anomaly detection
title_sort deep weakly-supervised anomaly detection
publisher Institutional Knowledge at Singapore Management University
publishDate 2023
url https://ink.library.smu.edu.sg/sis_research/8411
https://ink.library.smu.edu.sg/context/sis_research/article/9414/viewcontent/Deep_Weakly_supervised_Anomaly_Detection.pdf
_version_ 1787590771195510784