USAD: UnSupervised Anomaly Detection on Multivariate Time Series
Open accessThe DOI matches its Crossref record; 19 of the 20 references it cites were found in a registry or an index.
Abstract
The automatic supervision of IT systems is a current challenge at Orange. Given the size and complexity reached by its IT operations, the number of sensors needed to obtain measurements over time, used to infer normal and abnormal behaviors, has increased dramatically making traditional expert-based supervision methods slow or prone to errors. In this paper, we propose a fast and stable method called UnSupervised Anomaly Detection for multivariate time series (USAD) based on adversely trained autoencoders. Its autoencoder architecture makes it capable of learning in an unsupervised way. The use of adversarial training and its architecture allows it to isolate anomalies while providing fast training. We study the properties of our methods through experiments on five public datasets, thus demonstrating its robustness, training speed and high anomaly detection performance. Through a feasibility study using Orange's proprietary data we have been able to validate Orange's requirements on scalability, stability, robustness, training speed and high performance.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“It is beneficial to introduce an adversarial-based SSL, which can further amplify the difference between normal and anomalous data, such as USAD [66] and DAEMON [184].”
From Self-Supervised Learning for Time Series Analysis: Taxonomy, Progress, and Prospects (2024) · cites this paper as [66]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“The core idea behind USAD is to amplify the reconstruction error by adversarial training between two BAEs.”
From Self-Supervised Learning for Time Series Analysis: Taxonomy, Progress, and Prospects (2024) · cites this paper as [66]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“USAD [66] is a time series anomaly detection framework that includes two BAE models, and two BAE are defined as AE 1 and AE 2 , respectively.”
From Self-Supervised Learning for Time Series Analysis: Taxonomy, Progress, and Prospects (2024) · cites this paper as [66]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Adversarial-based regularization constraints can amplify the score difference between anomalies and normal points, thereby enhancing the accuracy of anomalies identification from another perspective, such as Minmax learning [66], [163] and adversarial learning [184].”
From Self-Supervised Learning for Time Series Analysis: Taxonomy, Progress, and Prospects (2024) · cites this paper as [66]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“RANSynCoders [63], USAD[66], AnomalyTrans [163], and DAEMON [184] employ autoencoder-based reconstruction SSL framework.”
From Self-Supervised Learning for Time Series Analysis: Taxonomy, Progress, and Prospects (2024) · cites this paper as [66]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Hochreiter, S., & Schmidhuber, J. (1997) · Neural Computation · doi:10.1162/neco.1997.9.8.1735 · Cited by 102,318 · cited 1×
“poral data, methods based on recurrent neural networks [7] (RNNs) are very popular.”
MethodGAN(Generative Adversarial Nets)
(2017) · Journal of Japan Society for Fuzzy Theory and Intelligent Informatics · doi:10.3156/jsoft.29.5_177_2 · Cited by 19,270
Learning Internal Representations by Error Propagation
Rumelhart, D. E., Hinton, G. E., & Williams, R. J. (1985) · doi:10.21236/ada164453 · Cited by 16,172
Liu, F. T., Ting, K. M., & Zhou, Z.-H. (2008) · 2008 Eighth IEEE International Conference on Data Mining · doi:10.1109/icdm.2008.17 · Cited by 6,601 · cited 1×
“Figure 2(C) presents the results for m ∈ [ 5 , 10 , 20 , 40 , 100 ] . dimension for Z causes a large loss of information at the encoding .”
Background5th International Conference on Learning Representations (ICLR 17)
Havrylov, S., & Titov, I. (2017) · Cited by 3,503
Robust Anomaly Detection for Multivariate Time Series through Stochastic Recurrent Neural Network
Su, Y., Zhao, Y., Niu, C., Liu, R., Sun, W., & Pei, D. (2019) · Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining · doi:10.1145/3292500.3330672 · Cited by 1,804 · cited 9×
“On top, the results obtained with SWaT and WADI datasets are presented, whereas the bottom part of the table reports obtained results from the benchmark proposed by [17], using three remaining datasets.”
BackgroundMethodResultInfluentialDetecting Spacecraft Anomalies Using LSTMs and Nonparametric Dynamic Thresholding
Hundman, K., Constantinou, V., Laporte, C., Colwell, I., & Soderstrom, T. (2018) · Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining · doi:10.1145/3219819.3219845 · Cited by 1,771
Schwartz, R., Dodge, J., Smith, N. A., & Etzioni, O. (2020) · Communications of the ACM · doi:10.1145/3381831 · Cited by 1,647
Deep Learning for Anomaly Detection: A Survey
Chalapathy, R., & Chawla, S. (2019) · arXiv · doi:10.48550/arxiv.1901.03407 · Cited by 1,240 · cited 1×
“Detecting unexpected behavior on a set of measurements correlated with each other over time is an active research discipline called anomaly detection in multivariate time series [2].”
BackgroundA Multimodal Anomaly Detector for Robot-Assisted Feeding Using an LSTM-Based Variational Autoencoder
Park, D., Hoshi, Y., & Kemp, C. C. (2018) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2018.2801475 · Cited by 1,115
Research works printed but not linked: 3
- Ian J. Goodfellow , Jean Pouget-Abadie , Mehdi Mirza , Bing Xu , David Warde-Farley , Sherjil Ozair , Aaron C. Courville , and Yoshua Bengio . 2014 . Generative Adversarial Nets. In Advances in Neural Information Processing Systems 27: Annual Conference on Neural Information Processing Systems 2014. Montreal, Quebec, Canada, 2672--2680. Ian J. Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron C. Courville, and Yoshua Bengio. 2014. Generative Adversarial Nets. In Advances in Neural Information Processing Systems 27: Annual Conference on Neural Information Processing Systems 2014. Montreal, Quebec, Canada, 2672--2680.
- Chuxu Zhang , Dongjin Song , Yuncong Chen , Xinyang Feng , Cristian Lumezanu , Wei Cheng , Jingchao Ni , Bo Zong , Haifeng Chen , and Nitesh V. Chawla . 2019. A Deep Neural Network for Unsupervised Anomaly Detection and Diagnosis in Multivariate Time Series Data . Proceedings of the AAAI Conference on Artificial Intelligence , Vol. 33 ( 2019 ), 1409--1416. Chuxu Zhang, Dongjin Song, Yuncong Chen, Xinyang Feng, Cristian Lumezanu, Wei Cheng, Jingchao Ni, Bo Zong, Haifeng Chen, and Nitesh V. Chawla. 2019. A Deep Neural Network for Unsupervised Anomaly Detection and Diagnosis in Multivariate Time Series Data. Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 33 (2019), 1409--1416.
- Bo Zong , Qi Song , Martin Renqiang Min , Wei Cheng , Cristian Lumezanu , Daeki Cho , and Haifeng Chen . 2018 . Deep autoencoding Gaussian mixture model for unsupervised anomaly detection . In 6th International Conference on Learning Representations, ICLR 2018. Toulon, France, 1--19. Bo Zong, Qi Song, Martin Renqiang Min, Wei Cheng, Cristian Lumezanu, Daeki Cho, and Haifeng Chen. 2018. Deep autoencoding Gaussian mixture model for unsupervised anomaly detection. In 6th International Conference on Learning Representations, ICLR 2018. Toulon, France, 1--19.
Cited by
Items 501–600 of 1,000
Brain-Inspired Real Time Anomaly Detection System for Mobile Networks
Ahmed, A. H., & Riegler, M. A. (2024) · IEEE International Conference on Communications workshops · doi:10.1109/iccworkshops59551.2024.10615603 · Cited by 3
“Unsupervised methods based on deep learning have recently received much attention [6] [7] [8] [9].”
BackgroundMethodInfluentialHuang, J., Xu, D., & Yang, T. (2024) · IEEE International Conference on Industrial Cyber-Physical Systems (ICPS) · doi:10.1109/icps59941.2024.10640001 · Cited by 3
“AE is generally trained by minimizing the reconstruction error of normal data, and then the error is used as the anomaly score to detect anomalies [23].”
MethodTian, F., Shi, X., Zhou, L., Chen, L., Ma, C., & Zhu, W. (2024) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn60899.2024.10650364 · Cited by 6
“In 2020, USAD [19] was proposed due to the large size and complexity of IT operations.”
BackgroundMethodUtilizing Duplicate Announcements for BGP Anomaly Detection
Verma, R. D., Keserwani, P. K., Jain, V. K., Govil, M. C., & Tilwari, V. (2025) · Telecom · doi:10.3390/telecom6010011 · Cited by 5
Fernando, D., Rodriguez, M. A., Arroba, P., Ismail, L., & Buyya, R. (2025) · ACM Transactions on Autonomous and Adaptive Systems · doi:10.1145/3725736 · Cited by 8
“The same AE model implementation mentioned in section IV-B was used for evaluations in this section as well.”
InfluentialResponsible Workload Sequence Anomaly Detection with Bi-View Feature Fusion
Liu, C., Wang, D., Li, D., Wang, K., & Zhang, J. (2025) · Journal of Network and Systems Management · doi:10.1007/s10922-025-09923-2 · Cited by 3
“…of Cyber-Physical Systems (CPSs), Nagarajan et al. [19] proposed an anomaly detection approach integrating CNN with a Kalman Filter based Gaussian Mixture Model (GMM) to identify and detect anomalous behavior in CPSs. Julien Audibert et al. [20] proposed a new GAN-based deep network model USAD.”
Liang, Q., & Yin, X. (2025) · Expert Systems · doi:10.1111/exsy.70102 · Cited by 4
“Recurrent neural networks, self-encoders, and Transformer networks were introduced to time series anomaly detection in papers like RobustTAD Gao et al. (2020), USAD Audibert et al. (2020), and TranAD Tuli et al. (2022).”
InfluentialGraph-Attention Diffusion for Enhanced Multivariate Time-Series Anomaly Detection
Lanko, V., & Makarov, I. (2024) · IEEE Open Journal of the Industrial Electronics Society · doi:10.1109/ojies.2024.3501014 · Cited by 5
“To tackle this problem modified versions of these metrics are usually adopted in time-series anomaly detection task [21].”
InfluentialWarping resilient robust anomaly detection for multivariate time series
Abilasha, S., & Bhadra, S. (2025) · Machine Learning · doi:10.1007/s10994-024-06689-7 · Cited by 3
“For each dataset, we have reported results with point adjustment (as in (Audibert et al., 2020; Li et al., 2021)) for WRADMts and all baselines.”
InfluentialWang, L., & Li, C. (2025) · ACM International Conference on Web Search and Data Mining (WSDM) · doi:10.1145/3701551.3703565 · Cited by 3
Schoch, F., Graf, P., Schmieg, T., Schloer, N., Wittenberg, C., Lanquillon, C., & Stache, N. C. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-93429-2_14 · Cited by 2
Deep Multimanifold Transformation-Based Multivariate Time Series Fault Detection
Liu, H., Qiu, X., Shi, Y., Xu, M., Zang, Z., & Lei, Z. (2025) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2025.3584988 · Cited by 4
“In time series anomaly detection, one-class classification (OCC) methods—such as USAD [4] and DAEMON [8]—typically assume access to purely normal data during training [5], [16].”
Real-Time Anomaly Detection of Electricity Time Series Data Based on Future-Guidance Network
Shi, Y., Xu, L., Wang, P., Wu, X., Wang, L., & Hou, Y. (2025) · IEEE Signal Processing Letters · doi:10.1109/lsp.2025.3598707 · Cited by 2
“These include the Multivariate Anomaly Detection with GAN (MAD-GAN) [19], the Transformer-based Anomaly Detection model (TranAD) [20], the Graph Deviation Network (GDN) [21], UnSupervised Anomaly Detection for multivariate time series (USAD) [22], and OmniAnomaly [15].”
KARMAD: KAN-Based Adversarial Robust Model for Anomaly Detection
Chen, F., Qiu, X., Ye, Y., Jing, R., Chen, Y., & Gao, D. (2025) · IEEE International Conference on Data Engineering (ICDE) · doi:10.1109/icde65448.2025.00256 · Cited by 2
“We compare KARMAD with state-of-the-art (SOTA) time series anomaly detection models in recent years, including DAGMM [40], DeepSVDD [30], OmniAnomaly [1], MS-CRED [41], USAD [14], THOC [31], GDN [33], CAE-M [42], TranAD [15], GRELEN [35], DCdetector [44], MT-GFlow [34], PeFAD [43] and TADNet [36].”
InfluentialAdversarial feature generation for ML-based intrusion detection in the petrochemical industry
Naqvi, S. S. A., Zhou, C., Xu, P., Li, Y., Jiashu, J., & Uzair, M. (2025) · Journal of Information Security and Applications · doi:10.1016/j.jisa.2025.104215 · Cited by 7
Distributed actor-critic method for multi-objective resource-aware neural architecture search
Tabrizchi, H., Razmara, J., & Javadi, B. (2025) · Applied Soft Computing · doi:10.1016/j.asoc.2025.114021 · Cited by 2
MADLLM: Multivariate Anomaly Detection via Pre-trained LLMs
Tao, W., Qu, X., Lü, K., Wan, J., Li, G., & Wang, J. (2025) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme59968.2025.11209795 · Cited by 6
“…that use different models, including classic machine learning methods (MERLIN [4]), traditional deep neural networks (OmniAnomaly [3], USAD [6], TimesNet [7]), attention mechanism-based models ( TranAD [10], ImDifussion [9]), and methods based on pre-trained LLMs (SFT [1], AnomalyLLM [13],…”
Real-Time Anomaly Detection in Industrial Cyber-Physical Systems Using Unsupervised Machine Learning
agure, Z. I., & Nggada, S. H. (2025) · International Symposium on Networks, Computers and Communications (ISNCC) · doi:10.1109/isncc66965.2025.11250502 · Cited by 3
Chen, L., Gao, X., Lu, H., Li, B., Xu, M., Feng, Y., Xue, B., & Wang, T. (2025) · Pattern Analysis and Applications · doi:10.1007/s10044-025-01580-0 · Cited by 3
Moon: A Modality Conversion-Based Efficient Multivariate Time Series Anomaly Detection
Yao, Y., Shi, Y., Chen, L., Fang, Z., Gao, Y., Leong Hou, U., Li, Y., & Li, T. (2025) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2025.3622154 · Cited by 3
“Index Terms —Anomaly detection, interpretable system, multivariate time series.”
Nivarthi, C. P., Huang, Z., Gruhl, C., & Sick, B. (2025) · Energy and AI · doi:10.1016/j.egyai.2025.100670 · Cited by 7
Ham, S., Jung, H., & Kim, E. Y. (2025) · IEEE Signal Processing Letters · doi:10.1109/lsp.2025.3619873 · Cited by 2
“We compare our method against state-of-the-art models from various paradigms: Transformer-based (TranAD [23], AT [24], SARAD [25], STTD [26]), MLP-based (USAD [27], COUTA [28], MMA [29]), graph-based (MT-Gflow [30]), diffusion-based (MODEM [31]), and CNN-based (TimesNet [32]).”
DUVET: Dual-View Enhanced Transformer for Anomaly Detection in Aviation Safety Monitoring
Li, C., Li, X., Chen, H., Zheng, L., Li, F., Zhao, X., Sun, H., Zhang, R., & Shang, J. (2026) · IEEE Sensors Journal · doi:10.1109/jsen.2025.3647934 · Cited by 3
“Recent advancements also feature approaches based on generative adversarial networks (GANs) [13], [14], the integration of graph neural networks for modeling both parametric and temporal dependencies (e.g., MTAD-GAT [15], and GDN [16]), as well as the use of transformers to further enhance temporal…”
Zurlo, G., Ronchieri, E., Cornali, R., Scarponi, L. B., Cesini, D., & dell'Agnello, L. (2026) · Journal of Business Analytics · doi:10.1080/2573234x.2025.2607576 · Cited by 6
A scale-adaptive spatio-temporal modeling approach for multivariate time-series anomaly detection
Chen, L., Wang, C., Lin, H., Chen, L., Jiang, J., & Lai, X. (2026) · Applied Intelligence · doi:10.1007/s10489-026-07120-5 · Cited by 2
“USAD[35] uses an adversarially trained autoencoder for fast and stable multivariate anomaly isolation, making it suitable for large-scale industrial monitoring.”
SCAT: A Spectral-Convolutional Anomaly Transformer for Multivariate Time Series Anomaly Detection
Zhang, S., Chen, S., & Li, J. (2026) · Electronics · doi:10.3390/electronics15030628 · Cited by 4
TSA-Net: Multivariate Time Series Anomaly Detection Based on Two-Stage Temporal Attention
Wu, H., Le, W., Jia, Z., Zhao, H., Zhang, S., & Zhang, Z.-S. (2026) · Sensors · doi:10.3390/s26031062 · Cited by 2
“• USAD [40]: Designed an encoder and two independent decoders to form two competing self-encoders.”
Multimodal Anomaly Detection for Microservice Systems via Grassmann Manifolds-Based Graph Fusion
He, S., Feng, K., Meng, K., Xie, K., & Zhao, X. (2026) · IEEE Transactions on Services Computing · doi:10.1109/tsc.2026.3665382 · Cited by 4
Adaptive Latent Distribution Modeling for Industrial Time Series Anomaly Detection
Liu, Y., Song, Y., Shu, S., Lin, F., Wang, J., & Guo, Y. (2026) · IEEE Transactions on Automation Science and Engineering · doi:10.1109/tase.2026.3674236 · Cited by 4
“These methods include DeepSAD [40], DeepSVDD [41], USAD [42], DAGMM [43], GANF [34], MTGFlow [10], FITS [44], and CrossAD [45].”
Shen, J., Zhang, H., Lin, W., Geng, Y., Li, J., Wang, J., & Xu, M. (2026) · IEEE Transactions on Networking · doi:10.1109/ton.2026.3680829 · Cited by 2
“First, unlike traditional time-series approaches [45], [47], Clover emphasizes that DoW detection is essentially a distribution-outlier problem .”
Multivariate time series anomaly detection using neighbor relation-guided Transformer
Zhang, J., Wu, Z., & Gan, H. (2026) · Pattern Recognition · doi:10.1016/j.patcog.2026.113835 · Cited by 2
Flexible O&M for Telecom Networks at Huawei: A Language Model-based Approach
Wu, S., Qin, Z., Wang, Z., Rao, S., Ye, Q., Quan, X., & Tian, G. (2021) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/3459637.3481919 · Cited by 4
“Previous researches mostly focus on detecting outliers in time-series [3, 30, 32, 38] and system logs [9, 14, 21].”
BackgroundVAE-TCN hybrid model for KPI Anomaly Detection
Wu, B., Xu, Q., Yao, Z., Tu, Y., & Chen, Y. (2022) · Asia-Pacific Network Operations and Management Symposium (APNOMS) · doi:10.23919/apnoms56106.2022.9919985 · Cited by 3
“We compare our model with seven state-of-the-art base-lines, including RNN-based methods: LSTM-VAE [10], Om-niAnomaly [11]; Graph-based method: MTAD-GAT [18]; CNN-based method: USAD [20]; Transformer-based methods: TransFram [23], AnomalyTran [24] and TranAD [28].”
BackgroundImproving Solar Flare Prediction by Time Series Outlier Detection
Wen, J., Islam, M. R., Ahmadzadeh, A., & Angryk, R. A. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-23480-4_13 · Cited by 3
Multi-stage Location for Root-Cause Metrics in Online Service Systems
Zhao, Y., Zhang, W., Wang, Z., Zhang, S., Sui, K., Dong, E., & Yang, J. (2023) · IEEE/IFIP Network Operations and Management Symposium (NOMS) · doi:10.1109/noms56928.2023.10154325 · Cited by 2
Anomaly Detection for Multivariate Telemetry Series of Satellite with Improved HVAE
Zhu, X., Pang, J., & Liang, X. (2023) · IEEE International Conference on Electronic Measurement & Instruments · doi:10.1109/icemi59194.2023.10270355 · Cited by 2
Zhang, X., Zhao, Y., Liu, C., Wang, L., Yang, X., Hou, Y., Lan, Z., Hu, X., Miao, B., Yang, M., et al. (2023) · Proceedings/Proceedings - International Symposium on Software Reliability Engineering · doi:10.1109/issre59848.2023.00084 · Cited by 2
Fernández, M. F., Mghames, S., & Bellotto, N. (2023) · Journal of Ambient Intelligence and Smart Environments · doi:10.3233/ais-230144 · Cited by 4
“An unsupervised anomaly detection for multivariate time series was proposed in [1] based on an adversarially-trained auto-encoder that outperformed previous unsupervised methods, such as Isolation Forests, Autoencoders, LSTM-VAE, DAGMM [62], and OmniAnomaly [48].”
MethodMulti-Scale Time Series Ensemble Learning for Information System Anomaly Detection
Wang, Y., Zhong, X., Shao, Y., Hu, J., Taishan, X., Bao, Y., & Li, W. (2023) · IEEE Smart World Congress (SWC) · doi:10.1109/swc57546.2023.10448910 · Cited by 2
“(5) USAD [23] is a method based on adversely trained au-toencoders which is able to isolate anomalies while providing fast training.”
BackgroundMethodSkip-Step Contrastive Predictive Coding for Time Series Anomaly Detection
Zhang, K., Wen, Q., Zhang, C., Sun, L., & Liu, Y. (2024) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp48485.2024.10447104 · Cited by 5
“USAD [13] proposes an encoder-decoder architecture within an adversarial training framework that allows it to isolate anomalies while providing fast training.”
BackgroundMethodUMAD-G: Unsupervised Multi-Modal Time Series Anomaly Detection via Graph
Lee, J., & Baek, J. (2024) · International Conference on Artificial Intelligence in Information and Communication (ICAIIC) · doi:10.1109/icaiic60209.2024.10463361 · Cited by 2
“We compare the performance of our proposed method with six popular anomaly detection models, including DAGMM [9], Omni-Anomaly [10], MAD-GAN [11], USAD [12], MTAD-GAT, and GDN.”
MethodMultidimensional Time Series Analysis for Anomaly Pattern Detection and Interpretation
Deng, Z., Kang, J., & Wang, X. (2024) · IEEE International Conference on Power, Electronics and Computer Applications (ICPECA) · doi:10.1109/icpeca60615.2024.10471100 · Cited by 2
Wu, Y.-X., & Dai, B.-R. (2024) · Lecture notes in computer science · doi:10.1007/978-981-97-2242-6_4 · Cited by 2
IAD-Net: Multivariate KPIs Interpretable Anomaly Detection with Dual Gated Residual Fusion Networks
Liu, W., Sun, D., Yang, H., Zhu, H., & Wang, Y. (2023) · IEEE International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom) · doi:10.1109/trustcom60117.2023.00103 · Cited by 4
“2) Metrics: Our evaluation primarily employs F1-score (hereafter, F1) to assess anomaly detection performance, following previous works [14], [16].”
MethodInfluentialAnomaly Detection in Multivariate Time Series with Contaminated Training Data Using VAE
Lv, D., Ma, M., Dong, H., Shang, Y., & Bao, K. (2024) · IEEE Conference on Big Data Security on Cloud (BigDataSecurity) · doi:10.1109/bigdatasecurity62737.2024.00010 · Cited by 3
“• USAD: USAD, presented in [8], is an anomaly detection model leveraging adversarially trained autoencoders.”
MethodUnsupervised MTS Anomaly Detection with Variational Autoencoders
Saravana, M. K., Roopa, M. S., Arunalatha, J. S., & Venugopal, K. R. (2024) · Lecture notes in networks and systems · doi:10.1007/978-981-97-2611-0_15 · Cited by 7
Anomaly Detection for Multivariate Time Series Based on Contrastive Learning and Autoformer
Shang, X., Zhang, J., Jiang, X., & Luo, H. (2024) · International Conference on Computer Supported Cooperative Work in Design (CSCWD) · doi:10.1109/cscwd61410.2024.10580672 · Cited by 2
“Existing methods generally perform worse on precision metrics than recall, such as [5], [6], [11], indicating that there are more FP in the detection results and poor generalization ability.”
BackgroundMethodSun, Q. Q., Li, Y., Hu, Z., Zhou, C., & Liu, L. (2024) · Lecture notes in computer science · doi:10.1007/978-981-97-5618-6_18 · Cited by 2
Obied, M. A., Zakaria, W., Ghaleb, F. F. M., Hassanien, A. E., & Abdelfattah, A. M. H. (2024) · CCF Transactions on Pervasive Computing and Interaction · doi:10.1007/s42486-024-00160-1 · Cited by 2
“USAD demonstrates robustness, training speed, and high performance compared with state-of-the-art methods on public datasets.”
InfluentialDetecting Cyber Attacks in Industrial Control Systems Using Duration-aware Representation Learning
Lan, B., & Yu, S. (2024) · International Conference on Computing and Artificial Intelligence (ICCAI) · doi:10.1145/3669754.3669812 · Cited by 2
“Audibert et al. [22] proposed an unsupervised adversarial training-based autoencoder combining GAN to amplify the reconstruction error for anomalous inputs.”
InfluentialCharacteristic-Aware Time-Series Representation Learning for Unsupervised Anomaly Detection
Yang, Y., Li, Z., Lin, P., Duan, J., Zhang, H., Yang, T., Huang, C., Lin, Z., Tong, Y., & Wang, Y. (2024) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn60899.2024.10650866 · Cited by 2
“The baselines include DAGMM [14], LSTM-VAE [15], MAD-GAN [1], OmniAnomaly [13], USAD [11], MTAD-GAT [10], GDN [12].”
BackgroundMethodTime-Series Representation Learning via Dual Reference Contrasting
Yu, R., Gong, Y., Wang, S., Si, J., Peng, X., Xu, B., & Lü, W. (2024) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/3627673.3679699 · Cited by 4
Alcacer, A., & Epifanio, I. (2024) · PLoS ONE · doi:10.1371/journal.pone.0311418 · Cited by 7
Shan, J., Cai, D., Fang, F., Khan, Z., & Fan, P. (2024) · IEEE Open Journal of the Communications Society · doi:10.1109/ojcoms.2024.3511951 · Cited by 4
“USAD [7] and TranAD [9] expanded the feature space between normal and abnormal by adversarial training between multiple decoders.”
InfluentialSowmya, K., & Ramesh, K. (2024) · Arabian Journal for Science and Engineering · doi:10.1007/s13369-024-09821-w · Cited by 5
Patch-Wise-Based Self-Supervised Learning for Anomaly Detection on Multivariate Time Series Data
Oh, S., Anh, L. H., Vu, D. T., Yu, G.-H., Hahn, M., & Kim, J. (2024) · Mathematics · doi:10.3390/math12243969 · Cited by 2
Pyramid Transformer for Multivariate Time Series Anomaly Detection in IoUT
Luo, Y., Chen, N., Huang, Z., Zheng, J., Qin, S., & Wu, D. O. (2024) · IEEE International Conference on Smart Internet of Things (SmartIoT) · doi:10.1109/smartiot62235.2024.00088 · Cited by 2
Self-Supervised Anomaly Detection Using Outliers for Multivariate Time Series
Hong, J., & Hur, Y. (2024) · IEEE Access · doi:10.1109/access.2024.3522325 · Cited by 3
He, Z., Chen, P., & Zheng, Z. (2025) · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3712195 · Cited by 2
“The best point-adjusted F1-score, which is widely used in previous literature [7, 14, 44, 53], is selected as an indicator of their efficacy here (more details are stated in Section 7.3).”
InfluentialUnsupervised Anomaly Detection for IoT Time Series Signals with GANs
Li, Y., Li, X., Lv, S., Chen, Y., Zhang, W., & Ding, Y. (2024) · IEEE International Conference on Signal, Information and Data Processing (ICSIDP) · doi:10.1109/icsidp62679.2024.10869232 · Cited by 4
Detecting Both Seen and Unseen Anomalies in Time Series
Liu, C., He, S., Li, S., Shi, Z., & Meng, W. (2025) · ACM Transactions on Knowledge Discovery from Data · doi:10.1145/3717071 · Cited by 5
“The ID features are expected to encompass adequate information that supports the recognition of seen anomalies [48] while the OOD features should exhibit a notable discrepancy between seen and unseen samples [3].”
Yu, D., Yao, M., Qi, P., & tao, D. (2025) · Communications in computer and information science · doi:10.1007/978-981-96-2189-7_9 · Cited by 2
Jia, X., Peng, W., Shen, C., Zhao, B., & Xun, P. (2025) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49660.2025.10888536 · Cited by 2
“The results are compared with the state-of-the-art methods including Anomaly Transformer(AT) [11], USAD [10], GANF [5], TranAD [12], PAD [17] and GST-Pro [6].”
DifNet: Difference-based multi-resolution decomposition for time series anomaly detection
Wang, H., Li, J., Chen, Y., Zou, X., Zeng, Z., Wu, J., Pan, C., Lu, Y., Gu, R., He, X., et al. (2025) · Applied Intelligence · doi:10.1007/s10489-025-06551-w · Cited by 2
Fährmann, D., Ihlefeld, M., Kuijper, A., & Damer, N. (2025) · IET Collaborative Intelligent Manufacturing · doi:10.1049/cim2.70032 · Cited by 6
Optimized edge weighting in graph neural networks for server performance anomaly detection
Zhang, H., Xu, H., Shi, J., Lin, X., Gao, Y., & Huang, Y. (2025) · Complex & Intelligent Systems · doi:10.1007/s40747-025-02069-3 · Cited by 3
“USAD [50]: Unsupervised Anomaly Detection (USAD) is anautoencoder-basedframeworkthatistrainedadversarially.”
Ha, T. W., Kang, J. M., & Kim, M. H. (2021) · Communications in computer and information science · doi:10.1007/978-3-030-92231-3_4 · Cited by 2
GUARD: Multigranularity-based Unsupervised Anomaly Detection Algorithm for Multivariate Time Series
Meng, F., Yang, Q., He, Z., Yang, S., & Tang, W. (2022) · IEEE International Conference on Cloud Computing and Intelligence Systems (CCIS) · doi:10.1109/ccis57298.2022.10016429 · Cited by 2
Maru, C., Brandherm, B., & Kobayashi, I. (2022) · IEEE International Conference on Big Data (BigData) · doi:10.1109/bigdata55660.2022.10020675 · Cited by 2
“It combines two transformers, learns by adversarial training, and uses reconstruction loss to detect anomalies as in USAD.”
BackgroundResultInfluentialShen, W., & Yang, G. (2022) · International Conference on Mechanical, Automation and Electrical Engineering (CMAEE) · doi:10.1109/cmaee58250.2022.00013 · Cited by 2
A Predictive Coding Approach to Multivariate Time Series Anomaly Detection
Qi, Z., Xie, H. L., & Shang, M. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-30637-2_13 · Cited by 2
Detecting unregistered users through semi-supervised anomaly detection with similarity datasets
Heo, D. H., Park, S. H., & Kang, S. J. (2023) · Journal Of Big Data · doi:10.1186/s40537-023-00791-8 · Cited by 2
Multivariate Time Series Anomaly Detection Method Based on mTranAD
Zhang, C., Li, Y., Li, J., Li, G., & Ma, H. (2023) · Lecture notes in computer science · doi:10.1007/978-981-99-4752-2_5 · Cited by 2
Effective Abnormal Activity Detection on Multivariate Time Series Healthcare Data
Niu, M., Zhao, Y., & Haddadi, H. (2023) · ACM Annual International Conference on Mobile Computing and Networking (MobiCom) · doi:10.1145/3570361.3615741 · Cited by 2
“[1], inputs of the prediction network can be original data as well as reconstruction results of the decoder.”
BackgroundTime Series Anomaly Detection with a Transformer Residual Autoencoder-Decoder
Wang, S., Wang, Y., & Li, W. (2023) · Lecture notes in computer science · doi:10.1007/978-981-99-8070-3_39 · Cited by 2
Efficiently Detecting Anomalies in IoT: A Novel Multi-Task Federated Learning Method
Hao, J., Chen, J., Chen, P., Wang, Y., Niu, X., Xu, L., & Xia, Y. (2024) · Lecture notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering · doi:10.1007/978-3-031-54531-3_6 · Cited by 2
Multivariate Time Series Anomaly Detection with Fourier Time Series Transformer
Ye, Y., He, Q., Zhang, P., Xiao, J., & Li, Z. (2023) · IEEE International Conference on Cloud Networking · doi:10.1109/cloudnet59005.2023.10490086 · Cited by 4
Explainable Anomaly Detection Using Masked Latent Generative ModelingPreprint
Lee, D.-S., Malacarne, S., & Aune, E. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4683591 · Cited by 2
Cyber-Attack Detection and Localization for SCADA system of CPSs
Li, D., Tang, J., Wu, S., Zheng, Z., & Ng, S.-K. (2025) · 2025 IEEE/ACM Second International Conference on AI Foundation Models and Software Engineering (Forge) · doi:10.1109/forge66646.2025.00038 · Cited by 1
APT Attack Detection of a New Power System based on DPI-transformer
Zhang, Y., & Li, Y. (2023) · Recent Advances in Electrical & Electronic Engineering (Formerly Recent Patents on Electrical & Electronic Engineering) · doi:10.2174/2352096516666230504111123 · Cited by 1
Fan, J., Si, Z., Wang, Z., Sun, D., Wu, J. X., & Wu, H. (2023) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn54540.2023.10191780 · Cited by 3
“2) Baselines: We chose four baselines for comparison: AE[31], LSTM-VAE [32], OmniAnomaly [6], and USAD[5].”
MethodInfluentialAnomaly Detection in Heterogeneous Time Series Data for Server-Monitoring Tasks
Yu, R., Xiao, F., Wang, Z., Yang, J., Han, D., Wang, Z., Jin, M., Li, C., Dong, E., & Xia, S. (2023) · IEEE Symposium on Computers and Communications (ISCC) · doi:10.1109/iscc58397.2023.10218155 · Cited by 1
Yu, P., Zhou, X., Chen, L., Qi, J., & Bai, L. (2023) · International Conference on Communication, Image and Signal Processing (CCISP) · doi:10.1109/ccisp59915.2023.10355803 · Cited by 1
“The MS-CRED[12] model, equipped with a multi-scale convective recursion Encoder-Decoder, displays exceptional competence in capturing intricate correlations and temporal dynamics across various modalities.”
MethodFault Detection Model Using Measurement Data in Fiber Optic Internet Lines
Battal, E., Özkan, G., Savaş, K., & Çakar, T. E. (2023) · International Informatics and Software Engineering Conference (IISEC) · doi:10.1109/iisec59749.2023.10391036 · Cited by 1
“While periodic faults can be detected using machine learning techniques such as STL decomposition, autoencoder and artificial neural network models are commonly used in non-periodic time series, particularly for large datasets, due to their relatively high performance [3].”
BackgroundMethodResultInfluentialCao, Y., Yan, J., Feng, H., Wang, T., Qiao, Y., & Peng, T. (2023) · China Automation Congress (CAC) · doi:10.1109/cac59555.2023.10451988 · Cited by 1
Feng, Y., Zhang, W., Sun, H., & Jiang, W. (2024) · Lecture notes in computer science · doi:10.1007/978-981-97-2242-6_1 · Cited by 1
MultiCon-AD: Integrated Contrastive Learning for Multivariate Time Series Anomaly Detection
Ma, Z., Xia, H., & Chen, G. (2024) · International Seminar on Artificial Intelligence, Networking and Information Technology (AINIT) · doi:10.1109/ainit61980.2024.10581787 · Cited by 1
“This technique circumvents the requirement for labeled datasets, thereby improving the efficiency of the detection process [4][10].”
MethodGraph Transformer: Anomaly Prediction and Interpretation in Multivariate Time Series
Li, J., Liu, C., & Ding, J. (2024) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn60899.2024.10650606 · Cited by 1
MTS anomaly detection based on temporal and feature correlation
Chen, S., Yan, D., & He, X. (2024) · Chinese Control Conference (CCC) · doi:10.23919/ccc63176.2024.10662146 · Cited by 1
“[1] However, as data grows exponentially in scale and complexity over time, this method becomes increasingly time-consuming.”
MethodSemi-Supervised Skeleton-Based Covert Cheating Detection in Electronic-Exams
Atabay, H. A., & Hassanpour, H. (2024) · Iranian Journal of Science and Technology Transactions of Electrical Engineering · doi:10.1007/s40998-024-00758-2 · Cited by 3
“The evaluation metrics are calculated over these score arrays evidenced by previous works (Malhotra et al. 2016; Audibert et al. 2020; Zhang et al. 2019).”
ChebyshevNet: a novel time series analysis model using Chebyshev polynomial
Diao, J., Cui, K., Huang, Y., Zhou, C., Li, J., & Song, H. (2024) · The Journal of Supercomputing · doi:10.1007/s11227-024-06672-y · Cited by 2
Multi-temporal-scale event detection and clustering in IoT systems
Park, Y., & Ko, I. (2024) · Internet of Things · doi:10.1016/j.iot.2024.101434 · Cited by 1
An enhanced time series anomaly detection model based on frequency domain analysis
Wang, J., & Jiang, Y. (2024) · International Conference on Big Data & Artificial Intelligence & Software Engineering (ICBASE) · doi:10.1109/icbase63199.2024.10762133 · Cited by 2
“We evaluated our proposed model's performance on test set, and compared it against six sota models for detecting anomalies in time series data, including DAGMM [17], MTAD-GAT [11], USAD [9], GDN [10], LSTM-NDT [16] and TimesNet [1].”
Fast POI anomaly detection using a weakly-supervised temporal state regression network
Yao, X. (2024) · Computational Urban Science · doi:10.1007/s43762-024-00151-z · Cited by 1
“USAD incorporates the encoder-decoder framework into adversarial training to gain detection stability (Audibert et al. 2020).”
PASTA : Neural Architecture Search for Anomaly Detection in Multivariate Time Series
Trirat, P., & Lee, J.-G. (2024) · IEEE Transactions on Emerging Topics in Computational Intelligence · doi:10.1109/tetci.2024.3508845 · Cited by 9
“The most well-known deep TSAD are prediction-based [6], [35], [36] and reconstruction-based [7], [8], [11], [12], [14], [15], [27], [35], [37], [38], [39], [40], [41], [42], [43], [44], [45] methods.”
USTG:Multivariate Time Series Anomaly Detection via Unsupervised Spatial-Temporal Graph Learning
Zhang, H., Wang, Y., & Han, Q. (2024) · International Conference on Behavioral, Economic, and Socio-Cultural Computing · doi:10.1109/besc64747.2024.10780512 · Cited by 2
Gao, T., Zuo, D., Shu, Y., Zhang, Z., Wen, D., & Qu, Y. (2024) · Lecture notes in computer science · doi:10.1007/978-981-96-0840-9_23 · Cited by 1
Shanmuganathan, V., & Suresh, A. (2025) · IETE Journal of Research · doi:10.1080/03772063.2024.2448588 · Cited by 2
“Currently, rules depend on anomaly detection, supervised anomaly detection, and unsupervised anomaly detection, which can be loosely classified into three groups [18,19].”
Lu, G., Lin, X. H., Pavlovski, M., Zhang, X., & Zhou, F. (2024) · IEEE International Conference on Big Data (BigData) · doi:10.1109/bigdata62323.2024.10825015 · Cited by 2
“Autoencoder-based methods [12]–[14] learn latent representations and detect anomalies based on reconstruction errors, while generative adversarial learning [15], [16] uses generators and discriminators to separate outliers through reconstruction errors.”
Time Series Anomaly Detection via Reconstruction-Limited Probability Estimation
Feng, Y., & Niu, Y. (2025) · Lecture notes in computer science · doi:10.1007/978-981-97-8702-9_30 · Cited by 2
Metadata
- Authors
- Audibert, J., Michiardi, P., Guyard, F., Marti, S., & Zuluaga, M. A.
- Affiliations
- EURECOM (FR)
- EURECOM, Biot, France
- Orange (France) (FR)
- Venue
- Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining
- Publisher
- ACM
- Published
- 2020-08-20
- Pages
- 3395-3404
- OpenAlex
- W3081497074
- Cited by
- 1,094
- References
- 20
- Citations by year
- 2026: 281 · 2025: 266 · 2024: 256 · 2023: 170 · 2022: 84 · 2021: 37
- FWCI
- 27.79
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) (publishedVersion, cc-by)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Anomaly Detection Techniques and Applications
- Physical Sciences › Computer Science › Computer Networks and Communications › Software System Performance and Reliability
- Physical Sciences › Computer Science › Computer Networks and Communications › Network Security and Intrusion Detection
- Keywords
- unsupervised anomaly detection, multivariate time series, adversarial autoencoder, network anomaly detection, orange, scalability, robustness
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.