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 901–1,000 of 1,000
Yu, J., Wang, X., Liu, H., & Meng, K. (2026) · Lecture notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering · doi:10.1007/978-3-032-23447-6_8
Gtr: graph-temporal relation transformer for multivariate time series anomaly detection
Shen, W., He, Z., & Li, S. (2026) · International Journal of Data Science and Analytics · doi:10.1007/s41060-026-01116-7
Abbasi, A., & Rodrigues, R. (2026) · IEEE International Conference on Industrial Technology (ICIT) · doi:10.1109/icit64854.2026.11490108
“3) Deep Generative Likelihood Models: OmniAnomaly [17] uses stochastic RNN–VAEs with normalizing flows to model complex distributions, while USAD [18] employs adversarially trained autoencoders for faster convergence.”
Li, J. W., Yu, Z., & Jiang, Q. (2026) · Lecture notes in electrical engineering · doi:10.1007/978-981-95-6553-5_22
An unsupervised multivariate time series anomaly detection method
Hu, J., Chen, W., Yang, Y., & Luo, Z. (2026) · Measurement · doi:10.1016/j.measurement.2026.121629 · Cited by 1
Kalro, P., Khan, R., & Jawale, A. (2026) · IEEE International Conference on AI Engineering and Innovations (AIEI) · doi:10.1109/aiei69164.2026.11497824
An Attention-Enhanced ConvNeXt Framework for Multivariate Anomaly Detection
Akter, M. S., Refah, Z. T., Hossain, M. D., Rashed, M. G., & Das, D. (2025) · International Conference on Computer and Information Technology (ICCIT) · doi:10.1109/iccit68739.2025.11491799 · Cited by 1
“Specifically, on SWaT, NPSR surpassing GDN, USAD [26], and TranAD [15].”
Mu, Y., Kong, Z., Zhang, M., Guo, H., & Wang, L. (2026) · Applied Intelligence · doi:10.1007/s10489-026-07248-4
卢芮伊, Yuan, W., Zhou, X., & Lan, S. (2026) · Lecture notes in computer science · doi:10.1007/978-981-95-7084-3_45
Zhang, Y., & Yin, X. (2026) · Expert Systems · doi:10.1111/exsy.70287
Chen, J., He, Q., Zhong, Y., Qin, X., & Wang, L. (2026) · Lecture notes in computer science · doi:10.1007/978-981-92-0372-7_36
MS-STNet: Multi-scale Spatio-Temporal Modeling for Multivariate Time Series Anomaly Detection
Wang, M., Li, J. F., Cui, Y., Li, F., Zhang, J. Z., & Wang, L. (2026) · Lecture notes in computer science · doi:10.1007/978-981-92-0372-7_37
Anomaly Detection in Connected Vehicle Sensor Data via Self-Supervised Deep Learning
Dahiya, A., Kumar, R., Gupta, M., & Tensingh, S. (2026) · International Conference on Inventive Computation Technologies (ICICT) · doi:10.1109/icict68280.2026.11511147
“Su et al. developed Omni Anomaly in 2019 which utilizes a stochastic recurrent neural network for the purpose of analyzing multivariate sensor data and achieved better results in benchmark tests [14].”
TFD-CAD: Collaborative time-frequency dual-branch framework for time series anomaly detection
Chen, J., Liu, P., Li, S. Y., Li, P., Li, X., & Hua, R. (2026) · Journal of King Saud University - Computer and Information Sciences · doi:10.1007/s44443-026-00805-4 · Cited by 1
Li, Z., Tong, J., & Liu, X. (2026) · IEEE Transactions on Dependable and Secure Computing · doi:10.1109/tdsc.2026.3694131 · Cited by 1
A Multi-modal Imbalance-Aware Method for Microservice Anomaly Detection
Tao, Y. (2026) · International Conference on Generative Artificial Intelligence and Information Security (GAIIS) · doi:10.1109/gaiis69281.2026.11519365
“Early studies on microservice anomaly detection mainly rely on a single data source, such as system metrics or execution traces [4] [5].”
Enhancing Unsupervised Anomaly Detection and Early Warning With Dual‐Attention LSTM ‐ AdvAE
Zhang, Z., Yu, J., & Weihua Yang, A. (2026) · Computational Intelligence · doi:10.1111/coin.70240 · Cited by 1
Wong, A. W., Chin, W. J., Tan, M. P., Mok, S.-Y., & Goh, C.-H. (2026) · Annals of Biomedical Engineering · doi:10.1007/s10439-026-04176-9
AI-Driven Multimodal Anomaly Detection Framework for Sensor-Based Cattle Health Monitoring
Swathi, M., Reddy, S. C. L., Shreya, S., Sumukh Anirudh, P., I, M. K. N., Ramesh, S., & Jeeva, S. (2026) · International Conference on Networking and Communications (ICNWC) · doi:10.1109/icnwc68145.2026.11518218
Chebyshev Low-Pass Attention Filter for Multivariate Time Series Anomaly Detection
Sun, P., Sun, D., Liu, X., & Han, L. (2026) · IEEE Internet of Things Journal · doi:10.1109/jiot.2026.3695127
“…The experiments compare CLAF against seven state-of-the-art baselines for sensor network anomaly detection: OmniAnomaly [33], MAD-GAN [37], USAD [38], ATUAD [20], Meta-MWDG [39], PeFAD [40] and FDTAD [41]. for each batch X ( ψ ) train in X train do 5: E = DataEmbedding ( X ( ψ ) train ) 6: 10:…”
Liao, X., Deng, W., Ma, H., & Mu, Y. (2026) · PLoS ONE · doi:10.1371/journal.pone.0349223
An Attack-Mechanism-Informed Informer for Sensor Deception Attack Detection in Process Industries
Chen, Y., Chen, M., Li, Y., & Shi, B. (2026) · Processes · doi:10.3390/pr14111686
“Unsupervised anomaly detection (USAD) formulates multivariate time-series anomaly detection through an unsupervised autoencoder-based framework [25].”
Liu, Y., Liu, X., Yao, J., Wang, Y., Guo, B., & Wang, Y. (2026) · Journal of Computer and Communications · doi:10.4236/jcc.2026.145002
ICAD: Rethinking the Role of Inference and Cues for the Anomaly Detection of Time Series in IIoT
Wu, Z., Zhu, L., Yin, Z., Zhang, S., Xu, X., Wei, X., & Yang, X. (2026) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2026.3682751
“USAD trainsautoencoderstoreconstructtheoriginalsamples[4].”
InfluentialJiang, K., Hao, X., Gao, B., Zheng, H., & Hu, Y. (2026) · International Journal of Pattern Recognition and Artificial Intelligence · doi:10.1142/s021800142650028x
Akhtar, N., & Rana, A. (2026) · International Conference on Augmented Intelligence and Sustainable Systems (ICAISS) · doi:10.1109/icaiss68683.2026.11525874
“Adversarial framework such as USAD further improved the separation of anomalies using double autoencoder training [20].”
Almufti, S. M. (2025) · Journal of Artificial Intelligence Machine Learning and Neural Network · doi:10.55529/jaimlnn.52.48.57
Solunke, M. H. B. (2025) · Journal of Artificial Intelligence Machine Learning and Neural Network · doi:10.55529/jaimlnn.51.137.150
Li, W., JIANG, J., Chen, Z., & Huang, H.-Z. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.6836289
Liu, F., Dai, Z., & Dou, H. (2026) · The Journal of Supercomputing · doi:10.1007/s11227-026-08613-3
Xu, J., Liu, H., & Xu, J. (2026) · Journal of big data and computing. · doi:10.62517/jbdc.202601123
Yao, Y., & Yang, Z.-X. (2026) · Advanced Engineering Informatics · doi:10.1016/j.aei.2026.104827
Asaad, R. R. (2025) · Journal of Artificial Intelligence Machine Learning and Neural Network · doi:10.55529/jaimlnn.52.79.87
Akanbi, B. M., Atumoshi, A. Y., Ishola, O. B., & Akanbi, A. O. (2026) · Federal University Gusau Faculty of Education Journal · doi:10.64348/zije.2026438 · Cited by 1
Chilukury, P. K., Reddy, S. R. B., Tirupati, K. K., & Methuku, N. K. (2026) · International Conference on Artificial Intelligence, Systems, and Emerging Technologies (ICAISET) · doi:10.1109/icaiset66439.2026.11541767
“The field then moved towards reconstruction-and density-based deep models, such as DAGMM, OmniAnomaly, MSCRED, DeepAnT, USAD and TranAD, which were able to better model multivariate temporal modeling but mostly were offline and static after deploying the model [9], [10], [11], [12], [13], [14].”
Cross-Modal Geometric Regularization for Multivariate Cybersecurity Anomaly Detection
Rai, S., Rodrigues, C., Czernichow, T., Lescos, D., & Nguyen, N. (2026) · IFIP advances in information and communication technology · doi:10.1007/978-3-032-27993-4_15
Towards AutoML Solutions for Predictive Maintenance
Papadopoulos, A., Giannoulidis, A., & Gounaris, A. (2026) · Procedia Computer Science · doi:10.1016/j.procs.2026.04.099
Bhoyar, D. M. R. (2026) · Journal of Artificial Intelligence Machine Learning and Neural Network · doi:10.55529/jaimlnn.61.31.41
Emergent detection of concept drift within the glia-inspired ‘rhythmic sharing’ algorithm
Whitehouse, I., Kang, H., & Losert, W. (2026) · npj Unconventional Computing · doi:10.1038/s44335-026-00067-3 · Cited by 1
Adaptive Unsupervised Anomaly Detection for Low-Quality Multivariate Time-Series Data
Liu, J., Yang, D., Wang, J., Gao, H., & Li, J. F. (2026) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2026.3700672
“USAD [22] proposed a reconstruction method based on adversarial learning, which trains two AE adversarially to amplify the reconstruction errors of inputs containing anomalies.”
DMSA: a dual-mask spatiatemporal autoencoder for multivariate time series anomaly detection
Bao, Y. (2026) · Applied Intelligence · doi:10.1007/s10489-026-07310-1
Nivarthi, C. P. (2026) · ACM Symposium on Applied Computing · doi:10.1145/3748522.3780036
SensitiveMamba: Selective State Space Modeling for Robust Time Series Anomaly Detection
Hur, M. G., Kim, D., & Min, M. (2026) · ACM Symposium on Applied Computing · doi:10.1145/3748522.3779727 · Cited by 1
Explainable Anomaly Detection in Financial Time Series: A Hybrid Deep Learning Approach
Hosseini, S. A., Grimaccia, F., Niccolai, A., Lorenzo, M., Casamatta, F., & Mousavi, A. (2025) · International Conference on Signal Processing and Intelligent Systems (ICSPIS) · doi:10.1109/icspis68676.2025.11551736
“Autoencoders, an instance of unsupervised neural networks intended to acquire efficient data representations, have proven to be particularly good instruments for anomaly identification [13], [14].”
Microservices Anomaly Detection Based on Adaptive Wavelet Convolutional Networks and Transformer
Tang, W., Jiang, C., Wang, G., & Liu, J. J. (2026) · IEEE Transactions on Services Computing · doi:10.1109/tsc.2026.3701934
Liu, C., Li, H., Zhao, S., Zeng, X., & Lu, X. B. (2026) · Entropy · doi:10.3390/e28060651 · Cited by 1
An Explainable Hybrid Framework for Process Fault Diagnosis
Gómez-González, P.-P., García-Fischer, A., Freire-Mahía, N., Michelena, Á., Jove, E., & ROLLE, J. L. C. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-29292-6_32
Meng, Z., & Xie, Z. (2026) · Lecture notes in computer science · doi:10.1007/978-981-95-4149-2_34
Gómez-González, P.-P., García-Fischer, A., Freire-Mahía, N., Michelena, Á., Jove, E., & ROLLE, J. L. C. (2026) · Communications in computer and information science · doi:10.1007/978-3-032-29254-4_6
Tu, J., Zhang, Z., Wu, J., & Wu, S. (2026) · Signal Image and Video Processing · doi:10.1007/s11760-026-05465-5
CTGAD: correlation-based transformer and GCN approach for cloud system anomaly detection
Liu, S., He, Q., Chen, Y., & Zhang, F. (2026) · Cluster Computing · doi:10.1007/s10586-026-06183-w
Guo, J., Kong, Y., Wang, J., Li, S., & Ren, L. (2026) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2026.3705659 · Cited by 1
“As for the anomaly detection model, it can employ any existing reconstruction-based MTS anomaly detection models, such as USAD [5], Anomaly Transformer [11] or ModernTCN [7].”
Generating Black-Box Adversarial Examples for Industrial Control Systems via Immune Co-Evolution
Huang, C., He, J., Li, W., Li, T., Ma, W., Fang, W., & Lan, X. (2026) · IEEE Internet of Things Journal · doi:10.1109/jiot.2026.3705384
“To address this gap, recent research has increasingly turned to machine learning (ML)-based anomaly detection [10], [11], [12], [13], [14].”
C-BADE: Concept-Based Anomaly Detection and Explanation Using Genetic Algorithm for Power Grids
Pokharel, D., & Li, Q. (2026) · ACM/IEEE International Conference on Cyber-Physical Systems (ICCPS) · doi:10.1109/iccps68698.2026.00023
“• USAD [15]: adversarially trained pair of autoencoders that improves reconstruction stability and sensitivity to rare anomalies.”
Zhang, Z., Wu, J., & Wu, S. (2026) · International Conference on Electronic Information Engineering, Big Data and Computer Technology (EIBDCT) · doi:10.1109/eibdct69742.2026.11566790
Zhang, J., Zhou, P., Jian-xi, Z., Chen, S., Li, Z., & Yang, L. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.6996733
Yazıcı, A., & Zengin, A. (2026) · Journal of Smart Systems Research · doi:10.58769/joinssr.1891109
Konkov, D., & Bozhko, L. (2026) · Intellectual Technologies on Transport · doi:10.20295/2413-2527-2026-246-82-90
ATOM: Adaptive Temporal Omniview Mixer for Lightweight Multivariate Time-Series Anomaly Detection
Jia, X., Peng, W., Xun, P., Zhao, B., & Shen, C. (2026) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2026.3702381
MAGIC-AD: A Lightweight Multi-Scale Anomaly Detection Framework for Industrial Control Systems
Shi, L., Zhang, J., Lei, X., & Yi, X. (2026) · International Conference on Robotics, Automation and Intelligent Transportation Systems (RAITS) · doi:10.1109/raits68656.2026.11580163
“Methods such as USAD [7] and MAD-GAN [8] incorporate adversarial training to enhance sensitivity, but they often suffer from training instability.”
Detecting CycleGAN-Spoofed AIS data using a GAN fingerprinting method with LSTM-Based classification
Zaman, S. M. A. U., Qamar, F., Mohd, M., Suhaimi, N. H. S., & Khandakar, A. (2026) · Frontiers in Future Transportation · doi:10.3389/ffutr.2026.1804269
A Weight-Optimized Federated Learning Framework for Leak Detection in Water Distribution Networks
Li, J., Feng, W., & Tang, X. (2026) · Water Resources Management · doi:10.1007/s11269-026-04812-y
“…a weight-optimized federated learning framework for leak detection in WDNs. USAD is used as the local anomaly-detection model at each client (Audibert et al. 2020), and a weight-optimization procedure is introduced at the aggregation stage to reduce the adverse effect of heterogeneous…”
Wojak-Strzelecka, N., Bobek, S., Pashami, S., & Nalepa, G. J. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7070753
Xu, G., Ye, Z., Huang, X., Liu, L., Wu, C., & Mumtaz, S. (2026) · IEEE Internet of Things Journal · doi:10.1109/jiot.2026.3710784
“Despite this progress, most earlier approaches mainly exploit anomalies through prediction or reconstruction residuals, and therefore treat abnormality as a deviation from learned normal patterns at the signal level [5], [6], [7], [23].”
InfluentialWicaksono, M., Imad, M., & Shin, S. Y. (2026) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2026.3711364
“More recently, the transformer-based models (TranADs) [40] and unsupervised learning frameworks [41] have been explored for multivariate time-series anomaly detection.”
Qiu, D. W., Zhou, H., Sun, Y., Wenyue, X., & Shang, L. (2026) · Lecture notes in computer science · doi:10.1007/978-981-92-2480-7_15
Azzari, A., Boaro, A., Basaldella, F., Nunes, S., Sala, F., Combi, C., Sala, P., & Bicego, M. (2026) · Journal of Healthcare Informatics Research · doi:10.1007/s41666-026-00245-9
Digital-Twin-Driven Anomaly Detection Using DQN in Industrial Control Systems
Wang, K., Zhou, C., Li, Y., Yang, M., & Yuan, L. (2026) · Lecture notes in computer science · doi:10.1007/978-981-92-2859-1_14
Wang, Y., Lu, J., Zhou, B., Zhang, Y., Li, X., Yuan, J., & Wang, L. (2026) · The Journal of Supercomputing · doi:10.1007/s11227-026-08690-4
A dual-stream deep learning architecture for business impact scoring and alert escalation
Javeed, M. S., Khatun, M. M., Alom, J., Islam, R., & Shoaib, H. A. (2026) · PLoS ONE · doi:10.1371/journal.pone.0350676 · Cited by 2
Yu, R., Yun, J., Jian, Z., Wu, X., Zhang, C., Wang, M., & Zhou, Z. (2026) · Lecture notes in computer science · doi:10.1007/978-981-92-3423-3_31
Lee, J., Yoon, T., Koo, W., & Kim, H. (2026) · IEEE Transactions on Automation Science and Engineering · doi:10.1109/tase.2026.3715199 · Cited by 1
“Unsupervised DNN-based MTAD methods are typically categorized into two classes: forecasting-based approaches [9], [27], [28], [29], [30] and reconstruction-based approaches [32], [33], [34], [35], [36], [37].”
Unsupervised Representation Learning for Time Series Anomaly Detection: Overview and new Trends
Hamida, O. I., Boudane, F., & Berrichi, A. (2026) · International Conference on Pattern Analysis and Intelligent Systems (PAIS) · doi:10.1109/pais69842.2026.11605406
“5) Hybrid Foundations: To combine the fast convergence of Autoencoders with the discriminative power of Generative Adversarial Networks , hybrid models such as USAD [6] have been proposed.”
SCT-5GC: Self-Conditioned Transformer for Mobile Core Network Telemetry Anomaly Detection
Issa, A., Kandil, N., & Hakem, N. (2026) · Canadian Conference on Electrical and Computer Engineering · doi:10.1109/ccece68150.2026.11610472
Chen, Y., Zhang, B., & Liu, C. (2026) · Measurement Science and Technology · doi:10.1088/1361-6501/ae8f62
Xu, Y., Xie, Z., He, Y. U. E. J. I., Zhu, Q., Zhang, Y., & Zhang, M. (2026) · IEEE Data Driven Control and Learning Systems Conference (DDCLS) · doi:10.1109/ddcls71227.2026.11610345
DS-STAE: Dual-Stream Spatio-Temporal Transformer for Context-Aware Anomaly Reconstruction in UAVs
Chen, Y., Cui, X., Liang, Z., & Tan, Z. (2026) · IEEE Data Driven Control and Learning Systems Conference (DDCLS) · doi:10.1109/ddcls71227.2026.11610517
Hierarchical multi-agent AI framework for cybersecurity in cyber–physical systems
Ferraro, A., Galli, A., Moscato, V., Orlando, G. M., & Russo, D. (2026) · Computer Networks · doi:10.1016/j.comnet.2026.112556
Xu, Y., Zhou, M.-Y., Luo, Y., He, Y. U. E. J. I., Zhu, Q., Zhang, Y., & Zhang, M. (2026) · IEEE Data Driven Control and Learning Systems Conference (DDCLS) · doi:10.1109/ddcls71227.2026.11610506
Knowledge-Constrained Multiscale Spatiotemporal Learning for Multivariate Anomaly Detection
Li, Y., Liu, Q., Fang, S., Tian, Y., Chen, F., & Zhou, C. (2026) · IEEE Data Driven Control and Learning Systems Conference (DDCLS) · doi:10.1109/ddcls71227.2026.11610240
“Representative baselines from diverse methods were selected, including graph-based dependency modeling (GDN [8], MGAT [16]), reconstruction/generative modeling (USAD [17], TranAD [11], VAE [18], IMDiff [19]), and self-supervised representation learning (CARLA [20]).”
One-Class LSTM Behavioral Modeling for Insider Threat Detection in Zero Trust Architectures
Robu, R. C., Marian, C. V., & Cioponea, A. P. (2026) · International Conference on Electronics, Computers and Artificial Intelligence (ECAI) · doi:10.1109/ecai69016.2026.11613700
“Multivariate time-series methods such as USAD [16] offer complementary unsupervised baselines for sequential behavioral data.”
GridCL for fine-grained load profiling in smart grids under limited labels
Zhang, L., Wang, J., Zhang, W., Li, K., Yao, D., Yang, B., Ke, X., Zhao, H., & Yao, Y. (2026) · PLoS ONE · doi:10.1371/journal.pone.0354752
Zhang, R., Liang, E., Hu, S., & Ke, Y. (2026) · International Journal on Data Science and Technology · doi:10.11648/j.ijdst.20261202.11
FastAD: Real-Time Multivariate Time Series Anomaly Detection via Structural Re-Parameterization
Wu, H., Le, W., Jia, Z., Zhang, Z., & Zhang, S. (2026) · International Conference on Computer and Communication Systems (ICCCS) · doi:10.1109/icccs69761.2026.11613452
“Specifically, we select recurrent and statistical models (LSTM-NDT [19], DAGMM [20]), reconstruction-based models (Om-niAnomaly [21], MAD-GAN [22], USAD [23], CAE-M [24]), graph-based approaches (MTAD-GAT [25], GDN [26]), and transformer-based baselines (TranAD [27], DTAAD [28]).”
Time series anomaly detection models with contrastive representation learning
Quijada, F., Drobnyh, K., & Runger, G. C. (2026) · Journal of Intelligent Information Systems · doi:10.1007/s10844-026-01078-z
“Particular to anomaly detection, vast research has focused in the development of unsupervised methods (Ahmad et al., 2017 ; Audibert et al., 2020; Chandola et al., 2009; Mejri et al., 2024), suitable for sparsely labeled data.”
Spatio-Temporal Multi-Sensor Anomaly Detection With Concurrent-Noise Distinction
Xiao, H., & Yuan, J. (2026) · IEEE Signal Processing Letters · doi:10.1109/lsp.2026.3718334
Kim, J. I., Lee, J., Kim, J., Heo, J., Jeong, J., & Kim, S. B. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-31666-0_15
Federated Learning for Multivariate Time Series Anomaly Detection in Industrial Automation
Nosrati, K., Uray, M., Messineo, S., Saßnick, O., & Huber, S. (2026) · Communications in computer and information science · doi:10.1007/978-3-032-34202-7_8
ASTER: Latent Pseudo-Anomaly Generation for Unsupervised Time-Series Anomaly Detection
Hermary, R., Hiçsönmez, S., Pineau, D., Shabayek, A. E. R., & Aouada, D. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-31654-7_22 · Cited by 2
A Hybrid Graph Variational Method for Anomaly Detection in Planetary Gearboxes
Li, Y., Yang, Q., Li, N., Lu, R., Cao, W., & Xing, L. (2026) · Youth Academic Annual Conference of Chinese Association of Automation (YAC) · doi:10.1109/yac71005.2026.11615449
Drift-Aware Memory-Augmented Spatio-Temporal Graph Attention for Industrial Anomaly Detection
Kim, Y., & Shin, D. (2026) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3770855.3818125
STLC-Net: A Short-Window Temporal Latent-Consistency Network for ADS-B Trajectory Anomaly Detection
Zhong, L., Liu, X., Xu, K., & Pan, W. (2026) · Aerospace · doi:10.3390/aerospace13080706
Wang, Y., Han, F., Li, K., Liu, S., Chen, T., & Peng, Z. (2026) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3770855.3817943
SEA-FGT: Frequency-Guided Transformer with Semantic Expert Augment for Time Series Anomaly Detection
Huang, W. X., Wang, Z., Huang, Y., Liu, J., & Ouyang, X. (2026) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3770855.3817823
Lee, J., Noh, J., Park, J., Ryu, H., Hanh, M., & Kim, J. (2026) · International Conference on Ubiquitous and Future Networks · doi:10.1109/icufn69619.2026.11628570
Adaptive Prototypical Contrastive Learning for Time Series Clustering
Li, W. (2026) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3770855.3817773
Wang, X., Wen, Y., He, Y., Xu, H., Zhao, Y., Haddad, M. F. C., & Ma, T. (2026) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3770855.3817978
Lu, Y., Gao, Y., Lan, T., Yang, Y., He, W., Wang, M., & Zhang, C. (2026) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3770855.3817912
A Time–Frequency Multi-Task Framework for Network Monitoring and Anomaly Detection
Qi, Q., Wang, C., Wang, X., He, B., Shang, J., Li, L., & Wang, J. (2026) · IEEE Transactions on Cognitive Communications and Networking · doi:10.1109/tccn.2026.3721519
“Twelve models were selected, including VAE [44], OmniAnomaly [45], USAD [46], MTAD-GAT [27], InterFusion [47], TranAD [48], Anomaly Transformer (Anom-alyTrans) [28], TimesNet [49], FPT [40], UAC-AD [29], CATCH [30], and TFMTNet.”
Duan, H., Zhang, Q., Ou, H., Li, C., Zhang, W., & Xu, Z. (2026) · Applied Energy · doi:10.1016/j.apenergy.2026.128639
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.