Adaptive memory networks with self-supervised learning for unsupervised anomaly detection
Unsupervised anomaly detection aims to build models to effectively detect unseen anomalies by only training on the normal data. Although previous reconstruction-based methods have made fruitful progress, their generalization ability is limited due to two critical challenges. First, the training data...
| Main Authors: | , , , , |
|---|---|
| Other Authors: | |
| Format: | Journal Article |
| Language: | English |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/179055 |
| _version_ | 1826112144027942912 |
|---|---|
| author | Zhang, Yuxin Wang, Jindong Chen, Yiqiang Yu, Han Qin, Tao |
| author2 | College of Computing and Data Science |
| author_facet | College of Computing and Data Science Zhang, Yuxin Wang, Jindong Chen, Yiqiang Yu, Han Qin, Tao |
| author_sort | Zhang, Yuxin |
| collection | NTU |
| description | Unsupervised anomaly detection aims to build models to effectively detect unseen anomalies by only training on the normal data. Although previous reconstruction-based methods have made fruitful progress, their generalization ability is limited due to two critical challenges. First, the training dataset only contains normal patterns, which limits the model generalization ability. Second, the feature representations learned by existing models often lack representativeness which hampers the ability to preserve the diversity of normal patterns. In this paper, we propose a novel approach called Adaptive Memory Network with Self-supervised Learning (AMSL) to address these challenges and enhance the generalization ability in unsupervised anomaly detection. Based on the convolutional autoencoder structure, AMSL incorporates a self-supervised learning module to learn general normal patterns and an adaptive memory fusion module to learn rich feature representations. Experiments on four public multivariate time series datasets demonstrate that AMSL significantly improves the performance compared to other state-of-the-art methods. Specifically, on the largest CAP sleep stage detection dataset with 900 million samples, AMSL outperforms the second-best baseline by 4 %+ in both accuracy and F1 score. Apart from the enhanced generalization ability, AMSL is also more robust against input noise. |
| first_indexed | 2024-10-01T03:02:08Z |
| format | Journal Article |
| id | ntu-10356/179055 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T03:02:08Z |
| publishDate | 2024 |
| record_format | dspace |
| spelling | ntu-10356/1790552024-07-17T08:30:48Z Adaptive memory networks with self-supervised learning for unsupervised anomaly detection Zhang, Yuxin Wang, Jindong Chen, Yiqiang Yu, Han Qin, Tao College of Computing and Data Science School of Computer Science and Engineering Computer and Information Science Artificial intelligence Unsupervised anomaly detection Unsupervised anomaly detection aims to build models to effectively detect unseen anomalies by only training on the normal data. Although previous reconstruction-based methods have made fruitful progress, their generalization ability is limited due to two critical challenges. First, the training dataset only contains normal patterns, which limits the model generalization ability. Second, the feature representations learned by existing models often lack representativeness which hampers the ability to preserve the diversity of normal patterns. In this paper, we propose a novel approach called Adaptive Memory Network with Self-supervised Learning (AMSL) to address these challenges and enhance the generalization ability in unsupervised anomaly detection. Based on the convolutional autoencoder structure, AMSL incorporates a self-supervised learning module to learn general normal patterns and an adaptive memory fusion module to learn rich feature representations. Experiments on four public multivariate time series datasets demonstrate that AMSL significantly improves the performance compared to other state-of-the-art methods. Specifically, on the largest CAP sleep stage detection dataset with 900 million samples, AMSL outperforms the second-best baseline by 4 %+ in both accuracy and F1 score. Apart from the enhanced generalization ability, AMSL is also more robust against input noise. Agency for Science, Technology and Research (A*STAR) AI Singapore Nanyang Technological University Submitted/Accepted version This work was supported in part by the National Key Research and Development Plan of China under Grant 2020YFC2007104, in part by the Natural Science Foundation of China under Grants 61972383, 61902377, and 61902379, in part by the Science and Technology Service Network Initiative, Chinese Academy of Sciences under Grant KFJ-STS-QYZD-2021-11-001, in part by the National Research Foundation, Singapore under its AI Singapore Programme under Grant AISG Award No. AISG2-RP-2020-019, in part by the RIE 2020 Advanced Manufacturing and Engineering (AME) Programmatic Fund under Grant A20G8b0102, Singapore, and the Nanyang Assistant Professorship (NAP). 2024-07-17T08:30:48Z 2024-07-17T08:30:48Z 2022 Journal Article Zhang, Y., Wang, J., Chen, Y., Yu, H. & Qin, T. (2022). Adaptive memory networks with self-supervised learning for unsupervised anomaly detection. IEEE Transactions On Knowledge and Data Engineering, 35(12), 12068-12080. https://dx.doi.org/10.1109/TKDE.2021.3139916 1041-4347 https://hdl.handle.net/10356/179055 10.1109/TKDE.2021.3139916 12 35 12068 12080 en AISG2-RP-2020-019 A20G8b0102 NAP IEEE Transactions on Knowledge and Data Engineering © 2022 IEEE. All rights reserved. This article may be downloaded for personal use only. Any other use requires prior permission of the copyright holder. The Version of Record is available online at http://doi.org/10.1109/TKDE.2021.3139916. application/pdf |
| spellingShingle | Computer and Information Science Artificial intelligence Unsupervised anomaly detection Zhang, Yuxin Wang, Jindong Chen, Yiqiang Yu, Han Qin, Tao Adaptive memory networks with self-supervised learning for unsupervised anomaly detection |
| title | Adaptive memory networks with self-supervised learning for unsupervised anomaly detection |
| title_full | Adaptive memory networks with self-supervised learning for unsupervised anomaly detection |
| title_fullStr | Adaptive memory networks with self-supervised learning for unsupervised anomaly detection |
| title_full_unstemmed | Adaptive memory networks with self-supervised learning for unsupervised anomaly detection |
| title_short | Adaptive memory networks with self-supervised learning for unsupervised anomaly detection |
| title_sort | adaptive memory networks with self supervised learning for unsupervised anomaly detection |
| topic | Computer and Information Science Artificial intelligence Unsupervised anomaly detection |
| url | https://hdl.handle.net/10356/179055 |
| work_keys_str_mv | AT zhangyuxin adaptivememorynetworkswithselfsupervisedlearningforunsupervisedanomalydetection AT wangjindong adaptivememorynetworkswithselfsupervisedlearningforunsupervisedanomalydetection AT chenyiqiang adaptivememorynetworkswithselfsupervisedlearningforunsupervisedanomalydetection AT yuhan adaptivememorynetworkswithselfsupervisedlearningforunsupervisedanomalydetection AT qintao adaptivememorynetworkswithselfsupervisedlearningforunsupervisedanomalydetection |