A Review on Outlier/Anomaly Detection in Time Series Data
Open accessThe DOI matches its Crossref record; 109 of the 109 references it cites were found in a registry or an index.
Abstract
Recent advances in technology have brought major breakthroughs in data collection, enabling a large amount of data to be gathered over time and thus generating time series. Mining this data has become an important task for researchers and practitioners in the past few years, including the detection of outliers or anomalies that may represent errors or events of interest. This review aims to provide a structured and comprehensive state-of-the-art on unsupervised outlier detection techniques in the context of time series. To this end, a taxonomy is presented based on the main aspects that characterize an outlier detection technique.
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.
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
Chandola, V., Banerjee, A., & Kumar, V. (2009) · ACM Computing Surveys · doi:10.1145/1541880.1541882 · Cited by 11,687
A survey on concept drift adaptation
Gama, J., Žliobaitė, I., Bifet, A., Pechenizkiy, M., & Bouchachia, A. (2014) · ACM Computing Surveys · doi:10.1145/2523813 · Cited by 3,735
A Survey of Outlier Detection Methodologies
Hodge, V., & Austin, J. (2004) · Artificial Intelligence Review · doi:10.1007/s10462-004-4304-y · Cited by 3,596
Hawkins, D. M. (1980) · doi:10.1007/978-94-015-3994-4 · Cited by 2,982
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 1×
“an outlier. Given that the results directly vary depending on that selection and few techniques give an automatic way to determine the threshold [Hundman et al. 2018;Lu et al. 2018;Siffer et al. 2017;Su et al. 2019], an interesting future line of research would be to deepen on the dynamic and adaptive selection of thr…”
BackgroundDetecting 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 · cited 7×
“theory ✓ ✓ ✓ S Keogh et al. [2002] Frequency ✗ ✓ ✓ S Wei et al. [2005] Similarity ✗ ✓ ✓ A Kumar et al.”
BackgroundMethodInfluentialDimensionality Reduction for Fast Similarity Search in Large Time Series Databases
Keogh, E., Chakrabarti, K., Pazzani, M., & Mehrotra, S. (2001) · Knowledge and Information Systems · doi:10.1007/pl00011669 · Cited by 1,573
A review on time series data mining
Fu, T.-C. (2011) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2010.09.007 · Cited by 1,500
Anomaly Detection Using Autoencoders with Nonlinear Dimensionality Reduction
Sakurada, M., & Yairi, T. (2014) · Proceedings of the MLSDA 2014 2nd Workshop on Machine Learning for Sensory Data Analysis · doi:10.1145/2689746.2689747 · Cited by 1,438 · cited 1×
“Sakurada and Yairi [2014] use this method, where the input of the autoencoder is a single multivariate point of the time series.”
MethodBest-practice recommendations for defining, identifying, and handling outliers
Aguinis, H., Gottfredson, R. K., & Joo, H. (2013) · Organizational Research Methods · doi:10.1177/1094428112470848 · Cited by 1,341
Outlier Detection for Temporal Data: A Survey
Gupta, M., Gao, J., Aggarwal, C. C., & Han, J. (2014) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2013.184 · Cited by 1,001 · cited 3×
“Although a number of surveys on outlier detection methods have been presented in the literature [Aggarwal 2016; Aguinis et al. 2013; Chandola et al. 2007, 2009; Hodge and Austin 2004; Xu et al. 2019], very few focus on temporal data, including time series [Gupta et al. 2014b].”
BackgroundEsling, P., & Agon, C. (2012) · ACM Computing Surveys · doi:10.1145/2379776.2379788 · Cited by 997 · cited 2×
“Alternatively, Galeano et al. (2006) suggest reducing the dimensionality with projection pursuit, which aims to find the best projections to identify outliers.”
BackgroundUnsupervised real-time anomaly detection for streaming data
Ahmad, S., Lavin, A., Purdy, S., & Agha, Z. (2017) · Neurocomputing · doi:10.1016/j.neucom.2017.04.070 · Cited by 995 · cited 9×
“(2017) Prediction 7 7 X X O/A Ahmad et al. (2017) Prediction 7 X X* X A Hochenbaum et al.”
BackgroundMethodInfluentialTime series shapelets: a new primitive for data mining
Ye, L., & Keogh, E. (2009) · Proceedings of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining · doi:10.1145/1557019.1557122 · Cited by 987 · cited 1×
“Other dissimilarity-based techniques are based on shapelets, which are representative time series subsequences [Ye and Keogh 2009].”
MethodThe problem of concept drift: definitions and related work
Tsymbal, A. (2004) · Cited by 929
Aggarwal, C. C. (2013) · Cited by 832
Joint Estimation of Model Parameters and Outlier Effects in Time Series
Chen, C., & Liu, L.-M. (1993) · Journal of the American Statistical Association · doi:10.1080/01621459.1993.10594321 · Cited by 777
Fox, A. J. (1972) · Journal of the Royal Statistical Society Series B: Statistical Methodology · doi:10.1111/j.2517-6161.1972.tb00912.x · Cited by 756
Outliers, level shifts, and variance changes in time series
Tsay, R. S. (1988) · Journal of Forecasting · doi:10.1002/for.3980070102 · Cited by 722
Anomaly Detection for Discrete Sequences: A Survey
Chandola, V., Banerjee, A., & Kumar, V. (2012) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2010.235 · Cited by 715
HOT SAX: Efficiently Finding the Most Unusual Time Series Subsequence
Keogh, E., Lin, J., & Fu, A. (2006) · Fifth IEEE International Conference on Data Mining (ICDM'05) · doi:10.1109/icdm.2005.79 · Cited by 696 · cited 3×
“…most straightforward idea consists of detecting the most unusual subsequence in a time series (denominated time series discord ) [Keogh et al. 2005; Lin et al. 2005], by comparing each subsequence with the others; that is, D where A is the set of all subsequences of X extracted by a sliding…”
BackgroundMethodDeepAnT: A Deep Learning Approach for Unsupervised Anomaly Detection in Time Series
Munir, M., Siddiqui, S. A., Dengel, A., & Ahmed, S. (2019) · IEEE Access · doi:10.1109/access.2018.2886457 · Cited by 696 · cited 4×
“In contrast, the CNN method described by Munir et al. [2019] is a prediction model-based Manuscript submitted to ACM technique, which is a direct extension of the method explained in Section 4.1 to detect time-aligned outliers that affect multiple variables.”
BackgroundMethodInfluentialAnomaly Detection in Streams with Extreme Value Theory
Siffer, A., Fouque, P.-A., Termier, A., & Largouet, C. (2017) · Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining · doi:10.1145/3097983.3098144 · Cited by 553 · cited 3×
“Additionally, they apply extreme value theory [Siffer et al. 2017] in the reference of normality to automatically select the threshold value.”
BackgroundMethodData stream clustering: A survey
Silva, J. A., Faria, E. R., Barros, R. C., Hruschka, E. R., Carvalho, A. C. P. L. F. D., & Gama, J. (2013) · ACM Computing Surveys · doi:10.1145/2522968.2522981 · Cited by 508
Aggarwal, C. C. (2013) · Springer eBooks · doi:10.1007/978-1-4614-6396-2 · Cited by 500
Clustering of time-series subsequences is meaningless: implications for previous and future research
Keogh, E., & Lin, J. (2005) · Knowledge and Information Systems · doi:10.1007/s10115-004-0172-7 · Cited by 464
Generic and Scalable Framework for Automated Time-series Anomaly Detection
Laptev, N., Amizadeh, S., & Flint, I. (2015) · Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining · doi:10.1145/2783258.2788611 · Cited by 425 · cited 2×
“[2008, 2009] Estimation Multivariate ✗ ✗ A Li et al. [2009] Dissimilarity Univariate ✓ ✓ O Sakurada and Yairi [2014] Estimation Multivariate ✗ ✗ A Zhou et al.”
BackgroundFinding surprising patterns in a time series database in linear time and space
Keogh, E., Lonardi, S., & Chiu, B. '.-C. (2002) · Proceedings of the eighth ACM SIGKDD international conference on Knowledge discovery and data mining · doi:10.1145/775047.775128 · Cited by 404 · cited 2×
“A paradigmatic example can be found in Keogh et al. [2002].”
MethodIncremental on-line learning: A review and comparison of state of the art algorithms
Losing, V., Hammer, B., & Wersing, H. (2018) · Neurocomputing · doi:10.1016/j.neucom.2017.06.084 · Cited by 382
Streaming Pattern Discovery in Multiple Time-Series
Papadimitriou, S., Sun, J., & Faloutsos, C. (2018) · doi:10.1184/r1/6609941.v1 · Cited by 354
Streaming Pattern Discovery in Multiple Time-Series
Papadimitriou, S., Sun, J., & Faloutsos, C. (2018) · doi:10.1184/r1/6609941 · Cited by 348
Anomaly detection in streaming environmental sensor data: A data-driven modeling approach
Hill, D. J., & Minsker, B. S. (2010) · Environmental Modelling & Software · doi:10.1016/j.envsoft.2009.08.010 · Cited by 300 · cited 1×
“Other methods use an autoregressive model [Hill and Minsker 2010] or an ARIMA model [Zhang et al. 2012], which obtain confidence intervals for the predictions instead of only point estimates.”
MethodAutomatic Anomaly Detection in the Cloud Via Statistical LearningPreprint
Hochenbaum, J., Vallis, O., & Kejariwal, A. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1704.07706 · Cited by 248 · cited 2×
“In contrast, in Hochenbaum et al. [2017], the ESD test explained in the previous paragraph is applied but the median and MAD are used for robustness instead of the mean and standard deviation.”
MethodDetecting distance-based outliers in streams of data
Angiulli, F., & Fassetti, F. (2007) · Proceedings of the sixteenth ACM conference on Conference on information and knowledge management · doi:10.1145/1321440.1321552 · Cited by 243 · cited 1×
“To take temporality into account, Angiulli and Fassetti [2007, 2010] and Ishimtsev et al. [2017] apply this method within a sliding window, which allows us to determine whether or not a new value of a streaming time series is an outlier upon arrival.”
MethodAutomatic outlier detection for time series: an application to sensor data
Basu, S., & Meckesheimer, M. (2007) · Knowledge and Information Systems · doi:10.1007/s10115-006-0026-6 · Cited by 235
Finding the most unusual time series subsequence: algorithms and applications
Keogh, E., Lin, J., Lee, S.-H., & Herle, H. V. (2006) · Knowledge and Information Systems · doi:10.1007/s10115-006-0034-6 · Cited by 228
Outlier Detection for Temporal Data
Gupta, M., Gao, J., Aggarwal, C. C., & Han, J. (2014) · Synthesis lectures on data mining and knowledge discovery · doi:10.2200/s00573ed1v01y201403dmk008 · Cited by 195
Outlier Detection for Multidimensional Time Series Using Deep Neural Networks
Kieu, T., Yang, B., & Jensen, C. S. (2018) · 2018 19th IEEE International Conference on Mobile Data Management (MDM) · doi:10.1109/mdm.2018.00029 · Cited by 195 · cited 1×
“Therefore, to account for temporal dependencies, Kieu et al. [2018] propose extracting features within overlapping sliding windows (e.g., statistical features or the difference between the Euclidean norms of two consecutive time windows) before applying the autoencoder.”
MethodLarge-Scale Unusual Time Series Detection
Hyndman, R. J., Wang, E., & Laptev, N. (2015) · 2015 IEEE International Conference on Data Mining Workshop (ICDMW) · doi:10.1109/icdmw.2015.104 · Cited by 189
Statistics-based outlier detection for wireless sensor networks
Zhang, Y., Hamm, N. A. S., Meratnia, N., Stein, A., van de Voort, M., & Havinga, P. J. M. (2012) · International Journal of Geographical Information Science · doi:10.1080/13658816.2012.654493 · Cited by 177
Outliers in multivariate time series
Tsay, R. S. (2000) · Biometrika · doi:10.1093/biomet/87.4.789 · Cited by 171
Finding anomalous periodic time series: An application to catalogs of periodic variable stars
Rebbapragada, U. D., Protopapas, P., Brodley, C. E., & Alcock, C. R. (2008) · Machine Learning · doi:10.1007/s10994-008-5093-3 · Cited by 166 · cited 1×
“Given that the well-known k -means clustering algorithm is not useful for unsynchronized variables because it uses the Euclidean distance, Rebbapragada et al. [2009] propose using Phased k -means in unsynchronized periodic multivariate time series.”
MethodTime-series Bitmaps: a Practical Visualization Tool for Working with Large Time Series Databases
Kumar, N., Lolla, V. N., Keogh, E., Lonardi, S., Ratanamahatana, C. A., & Wei, L. (2005) · Proceedings of the 2005 SIAM International Conference on Data Mining · doi:10.1137/1.9781611972757.55 · Cited by 164 · cited 2×
“Wei et al. [2005] and Kumar et al. [2005] use this reference of normality to obtain ˆ S .”
BackgroundMethodApproximations to Magic: Finding Unusual Medical Time Series
Lin, J., Keogh, E., Fu, A. W.-C., & Herle, H. M.-L. V. (2005) · Proceedings - IEEE Symposium on Computer-Based Medical Systems · doi:10.1109/cbms.2005.34 · Cited by 140 · cited 3×
“…most straightforward idea consists of detecting the most unusual subsequence in a time series (denominated time series discord ) [Keogh et al. 2005; Lin et al. 2005], by comparing each subsequence with the others; that is, D where A is the set of all subsequences of X extracted by a sliding…”
BackgroundMethodTemporal Outlier Detection in Vehicle Traffic Data
Li, X., Li, Z., Han, J., & Lee, J.-G. (2009) · IEEE International Conference on Data Engineering (ICDE) · doi:10.1109/icde.2009.230 · Cited by 138
Ratanamahatana, C. A., Lin, J., Gunopulos, D., Keogh, E., Vlachos, M., & Das, G. (2009) · Springer eBooks · doi:10.1007/978-0-387-09823-4_56 · Cited by 129
Detection and Characterization of Anomalies in Multivariate Time Series
Cheng, H., Tan, P., Potter, C. S., & Klooster, S. A. (2009) · SIAM International Conference on Data Mining · doi:10.1137/1.9781611972795.36 · Cited by 127
Outlier Detection in Multivariate Time Series by Projection Pursuit
Galeano, P., Peña, D., & Tsay, R. S. (2006) · Journal of the American Statistical Association · doi:10.1198/016214505000001131 · Cited by 126
Automated Load Curve Data Cleansing in Power Systems
Chen, J., Li, W., Lau, A., Cao, J., & Wang, K. (2010) · IEEE Transactions on Smart Grid · doi:10.1109/tsg.2010.2053052 · Cited by 123
Recent Progress of Anomaly Detection
Xu, X., Liu, H., & Yao, M. (2019) · Complexity · doi:10.1155/2019/2686378 · Cited by 122 · cited 2×
“Although a number of surveys on outlier detection methods have been presented in the literature [Aggarwal 2016; Aguinis et al. 2013; Chandola et al. 2007, 2009; Hodge and Austin 2004; Xu et al. 2019], very few focus on temporal data, including time series [Gupta et al. 2014b].”
BackgroundEfficient Periodicity Mining in Time Series Databases Using Suffix Trees
Rasheed, F., Alshalalfa, M., & Alhajj, R. (2010) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2010.76 · Cited by 121 · cited 1×
“The algorithm checks the periodicity based on the algorithm described in Rasheed et al. [2011]. As shown in Fig.”
MethodDetecting heat events in dairy cows using accelerometers and unsupervised learning
Shahriar, M. S., Smith, D., Rahman, A., Freeman, M., Hills, J. L., Rawnsley, R., Henry, D., & Bishop-Hurley, G. J. (2016) · Computers and Electronics in Agriculture · doi:10.1016/j.compag.2016.08.009 · Cited by 120
Assumption-free anomaly detection in time series
Li, W., Kumar, N. A., Lolla, V. S. S. C., Keogh, E., Lonardi, S., & Ratanamahatana, C. A. (2005) · Cited by 109
Song, S., Zhang, A., Wang, J., & Yu, P. S. (2015) · ACM SIGMOD International Conference on Management of Data (SIGMOD) · doi:10.1145/2723372.2723730 · Cited by 109 · cited 2×
“…such as B-splines or kernels [Chen et al. 2010] or variants of the Exponentially Weighted Moving Average (EWMA) method [Carter and Streilein 2012], Song et al. [2015] and Zhang et al. [2016] use slope constraints, Mehrang et al. [2015] assume that the data are approximately normal if the outliers…”
BackgroundMethodWAT: Finding Top-K Discords in Time Series Database
Bu, Y., Leung, T.-W., Fu, A. W.-C., Keogh, E., Pei, J., & Meshkin, S. (2007) · SIAM International Conference on Data Mining · doi:10.1137/1.9781611972771.43 · Cited by 107 · cited 2×
“…both the heuristic reordering and the pruning for discord detection, which may be based on the Haar wavelet transformation and augmented trie [Bu et al. 2007; Fu et al. 2006], on bit representation clustering [Li et al. 2013], on bounding boxes [Sanchez and Bustos 2014], on clustering…”
MethodCarreño, A., Inza, I., & Lozano, J. A. (2019) · Artificial Intelligence Review · doi:10.1007/s10462-019-09771-y · Cited by 99
Self-citationDistance-based outlier queries in data streams: the novel task and algorithms
Angiulli, F., & Fassetti, F. (2010) · Data Mining and Knowledge Discovery · doi:10.1007/s10618-009-0159-9 · Cited by 98
Mining Deviants in a Time Series Database
Jagadish, H. V., Koudas, N., & Muthukrishnan, S. (1999) · Very Large Data Bases · Cited by 97
Anomaly detection in time series data using a fuzzy c-means clustering
Izakian, H., & Witold Pedrycz, C. (2013) · Joint IFSA World Congress and NAFIPS International Conference · doi:10.1109/ifsa-nafips.2013.6608627 · Cited by 96 · cited 3×
“Alternatively, Izakian and Pedrycz [2013] employ Fuzzy C-Means clustering on the raw data, allowing each subsequence to belong to more than one cluster. In particular, the authors use the Euclidean distance in Equation (7). Note that these two clustering approaches are not applied to streaming time series in the origi…”
MethodInfluentialProbabilistic anomaly detection in natural gas time series data
Akouemo, H. N., & Povinelli, R. J. (2015) · International Journal of Forecasting · doi:10.1016/j.ijforecast.2015.06.001 · Cited by 93 · cited 2×
“[2016] Estimation ✗ ✓ ✓ ✓ Dirty Akouemo and Povinelli [2016] Estimation ✓ ✓ ✓ ✓ O/A Xu et al.”
Yang, J., Wang, W. W., & Yu, P. S. (2001) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/502512.502571 · Cited by 91 · cited 1×
“Periodic subsequence outliers are unusual subsequences that repeat over time [Yang et al. 2001; Rasheed and Alhajj 2014].”
BackgroundOutlier Detection Using Random Walks
Moonesinghe, H. D. K., & Tan, P. (2006) · IEEE International Conference on Tools with Artificial Intelligence (ICTAI) · doi:10.1109/ictai.2006.94 · Cited by 83 · cited 1×
“To this end, a Markov chain random walk model is applied in a graph where each node is a subsequence and each edge the pairwise similarity value between the nodes (Moonesinghe and Tan, 2006).”
MethodUnsupervised outlier detection for time series by entropy and dynamic time warping
Benkabou, S.-E., Benabdeslem, K., & Canitia, B. (2017) · Knowledge and Information Systems · doi:10.1007/s10115-017-1067-8 · Cited by 79 · cited 1×
“Even so, the process can be sped-up by reordering the search using heuristics and pruning off some fruitless calculations with the HOT-SAX algorithm [Keogh et al. 2005, 2007; Lin et al. 2005], which is based on the SAX discrete representation.”
MethodSimilarity Measure Selection for Clustering Time Series Databases
Mori, U., Mendiburu, A., & Lozano, J. A. (2015) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2015.2462369 · Cited by 74 · cited 1×
“For example, Papadimitriou et al. [2005] propose an incremental Principal Component Analysis (PCA) algorithm to determine the new independent variables.”
MethodSelf-citationSequential Data Cleaning: A Statistical Approach
Zhang, A., Song, S., & Wang, J. (2016) · ACM SIGMOD International Conference on Management of Data (SIGMOD) · doi:10.1145/2882903.2915233 · Cited by 74 · cited 2×
“…[Chen et al. 2010] or variants of the Exponentially Weighted Moving Average (EWMA) method [Carter and Streilein 2012], Song et al. [2015] and Zhang et al. [2016] use slope constraints, Mehrang et al. [2015] assume that the data are approximately normal if the outliers are not taken into…”
BackgroundMethodHu, M. C., Feng, X., Ji, Z., Yan, K., & Zhou, S. (2018) · Information Sciences · doi:10.1016/j.ins.2018.10.047 · Cited by 74
Finding Time Series Discords Based on Haar Transform
Fu, A. W.-C., Leung, O. T.-W., Keogh, E., & Lin, J. (2006) · Lecture notes in computer science · doi:10.1007/11811305_3 · Cited by 70 · cited 1×
“…reordering and the pruning for discord detection, which may be based on the Haar wavelet transformation and augmented trie [Bu et al. 2007; Fu et al. 2006], on bit representation clustering [Li et al. 2013], on bounding boxes [Sanchez and Bustos 2014], on clustering categorical data [Chau…”
Data Improving in Time Series Using ARX and ANN Models
Akouemo, H. N., & Povinelli, R. J. (2017) · IEEE Transactions on Power Systems · doi:10.1109/tpwrs.2017.2656939 · Cited by 70
GrammarViz 3.0: Interactive Discovery of Variable-Length Time Series Patterns
Senin, P., Lin, J., Wang, X., Oates, T., Gandhi, S., Boedihardjo, A. P., Chen, C., & Frankenstein, S. (2018) · ACM Transactions on Knowledge Discovery from Data · doi:10.1145/3051126 · Cited by 68
Time series outlier detection and imputation
Akouemo, H. N., & Povinelli, R. J. (2014) · IEEE Power Engineering Society General Meeting · doi:10.1109/pesgm.2014.6939802 · Cited by 61 · cited 6×
“[2017] use the STL decomposition, and Akouemo and Povinelli [2014, 2016, 2017] use ARIMA models with exogenous inputs, linear regression or Artificial Neural Networks (ANNs). Although most of these models can also be used in prediction, in this case, the outliers are detected in the residual set using past and future…”
MethodInfluentialA Framework for Periodic Outlier Pattern Detection in Time-Series Sequences
Rasheed, F., & Alhajj, R. (2013) · IEEE Transactions on Cybernetics · doi:10.1109/tsmcc.2013.2261984 · Cited by 59 · cited 1×
“, Senin et al. [2015]; Wang et al.”
BackgroundThe user side of sustainability: Modeling behavior and energy usage in the home
Chen, C., Cook, D. J., & Crandall, A. S. (2012) · Pervasive and Mobile Computing · doi:10.1016/j.pmcj.2012.10.004 · Cited by 58
Probabilistic reasoning for streaming anomaly detection
Carter, K. M., & Streilein, W. (2012) · IEEE Statistical Signal Processing Workshop (SSP) · doi:10.1109/ssp.2012.6319708 · Cited by 57 · cited 1×
“A detailed overview of the existing research regarding outlier detection in discrete sequences can be found in Chandola et al. [2012], which highlights the applicability of those techniques to time series data.”
BackgroundOutliers Detection in Multivariate Time Series by Independent Component Analysis
Baragona, R., & Battaglia, F. (2007) · Neural Computation · doi:10.1162/neco.2007.19.7.1962 · Cited by 55 · cited 2×
“Similarly, Baragona and Battaglia [2007] propose using Independent Component Analysis (ICA) to obtain a set of unobservable independent nonGaussian variables, assuming that the observed data has been generated by a linear combination of those variables.”
MethodUsing Gaussian Mixture Models to Detect Outliers in Seasonal Univariate Network Traffic
Reddy, A., Ordway-West, M., Lee, M., Dugan, M., Whitney, J., Kahana, R., Ford, B., Muedsam, J., Henslee, A., & Rao, M. (2017) · IEEE Security and Privacy Workshops (SPW) · doi:10.1109/spw.2017.9 · Cited by 54 · cited 2×
“…Moving Average (EWMA) method [Carter and Streilein 2012], Song et al. [2015] and Zhang et al. [2016] use slope constraints, Mehrang et al. [2015] assume that the data are approximately normal if the outliers are not taken into account, and Reddy et al. [2017] use Gaussian Mixture Models (GMM).”
MethodAn Outlier Detection Algorithm Based on Cross-Correlation Analysis for Time Series Dataset
Lu, H., Liu, Y., Fei, Z., & Guan, C. (2018) · IEEE Access · doi:10.1109/access.2018.2870151 · Cited by 54 · cited 3×
“Conversely, Manuscript submitted to ACM Shahriar et al. [2016] and Lu et al. [2018] use methods that include temporality in the transformation phase.”
MethodInfluentialExact variable-length anomaly detection algorithm for univariate and multivariate time series
Wang, X., Lin, J., Patel, N. S., & Braun, M. (2018) · Data Mining and Knowledge Discovery · doi:10.1007/s10618-018-0569-7 · Cited by 53
Finding time series discord based on bit representation clustering
Li, G., Bräysy, O., Jiang, L., Wu, Z., & Wang, Y. (2013) · Knowledge-Based Systems · doi:10.1016/j.knosys.2013.09.015 · Cited by 52 · cited 2×
“heuristic reordering and the pruning for discord detection, which may be based on the Haar wavelet transformation and augmented trie [Bu et al. 2007;Fu et al. 2006], on bit representation clustering [Li et al. 2013], on bounding boxes [Sanchez and Bustos2014], on clustering categorical data [Chau et al. 2018], and on…”
BackgroundMethodMining deviants in time series data streams
Muthukrishnan, S., Shah, R. R., & Vitter, J. S. (2004) · doi:10.1109/ssdbm.2004.51 · Cited by 52 · cited 2×
“Some years later Muthukrishnan et al. [2004] observed that for any bucket, the optimal set of k = |D | deviants always consists of the l highest and remaining k − l lowest values within the bucket, for some l ≤ k .”
BackgroundTime series anomaly detection based on shapelet learning
Beggel, L., Kausler, B. X., Schiegg, M., Pfeiffer, M., & Bischl, B. (2018) · Computational Statistics · doi:10.1007/s00180-018-0824-9 · Cited by 52
Online anomaly detection for long-term ECG monitoring using wearable devices
Carrera, D., Rossi, B., Fragneto, P., & Boracchi, G. (2018) · Pattern Recognition · doi:10.1016/j.patcog.2018.11.019 · Cited by 52
A Piecewise Aggregate pattern representation approach for anomaly detection in time series
Ren, H., Liu, M., Li, Z., & Witold Pedrycz, C. (2017) · Knowledge-Based Systems · doi:10.1016/j.knosys.2017.07.021 · Cited by 45 · cited 2×
“Additionally, apart from discretization, raw data has also been used directly to obtain representations based on dictionaries (e.g., Carrera et al. [2016]), exemplars (e.g., Jones et al. [2016]), or connectivity values (e.g., Ren et al. [2017]).”
BackgroundMethodConformal k-NN Anomaly Detector for Univariate Data StreamsPreprint
Ishimtsev, V., Nazarov, I., Bernstein, A., & Burnaev, E. V. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1706.03412 · Cited by 39
Exemplar learning for extremely efficient anomaly detection in real-valued time series
Jones, M., Nikovski, D. N., Imamura, M., & Hirata, T. (2016) · Data Mining and Knowledge Discovery · doi:10.1007/s10618-015-0449-3 · Cited by 38 · cited 2×
“, Keogh et al. [2002]), equal-width binning (e.g., Chen and Cook [2011]; Chen et al. [2013]; Rasheed and Alhajj [2014]), or SAX (e.g., Senin et al. [2015]; Wang et al. [2018]). These discretization techniques can also be used as a starting point to obtain other representations, such as bitmaps (e.g., Kumar et al. [200…”
BackgroundMethodZhou, Y., Qin, R., Xu, H., Sadiq, S. W., & Yu, Y. F. (2018) · Sensors · doi:10.3390/s18082628 · Cited by 37
Cheng, H., Tan, P., Potter, C. S., & Klooster, S. A. (2008) · IEEE ... International Conference on Data Mining workshops · doi:10.1109/icdmw.2008.48 · Cited by 35 · cited 3×
“Thus, possible future work may include temporal information to techniques that do not consider it (e.g., Cheng et al. [2008, 2009]; Siffer et al. [2017]) to see if the results improve.”
BackgroundMethodInfluentialClustering of time-series subsequences is meaningless
KeoghEamonn, & LinJessica (2005) · Knowledge and Information Systems · doi:10.5555/2993947.2993982 · Cited by 35
Nonparametric Event Detection in Multiple Time Series for Power Distribution Networks
Zhou, Y., Arghandeh, R., Zou, H., & Spanos, C. J. (2018) · IEEE Transactions on Industrial Electronics · doi:10.1109/tie.2018.2840508 · Cited by 33
Energy outlier detection in smart environments
Chen, C., & Cook, D. J. (2011) · Cited by 32
Outlier detection in weight time series of connected scales
Mehrang, S., Helander, E. E., Pavel, M., Chieh, A. Y., & Korhonen, I. (2015) · IEEE International Conference on Bioinformatics and Biomedicine (BIBM) · doi:10.1109/bibm.2015.7359896 · Cited by 32 · cited 2×
“The most simple estimation models are based on constant or piecewise constant models, where basic statistics such as the median [Basu and Meckesheimer 2007] or the Median Absolute Deviation (MAD) [Mehrang et al. 2015] are used to obtain the expected value ˆ x t .”
MethodMining Surprising Periodic Patterns
Yang, J., Wang, W., & Yu, P. S. (2004) · Data Mining and Knowledge Discovery · doi:10.1023/b:dami.0000031631.84034.af · Cited by 30
Mining deviants in time series data streams
Muthukrishnan, S., Shah, R., & Vitter, J. S. (2004) · International Conference on Scientific and Statistical Database Management · doi:10.1109/ssdm.2004.1311192 · Cited by 29
Adaptive Threshold for Anomaly Detection Using Time Series Segmentation
Dani, M. C., Jollois, F.-X., Nadif, M., & Freixo, C. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-26555-1_10 · Cited by 26 · cited 1×
“Dani et al. [2015] use the mean of each segment to determine the expected value of the points within that segment.”
BackgroundHolešovský, J., Čampulová, M., & Michálek, J. (2017) · Atmospheric Pollution Research · doi:10.1016/j.apr.2017.06.005 · Cited by 26 · cited 1×
“For instance, some authors model the structure of the data using smoothing methods such as B-splines or kernels [Chen et al. 2010; Holešovský et al. 2018; Čampulová et al. 2018] or variants of the Exponentially Weighted Moving Average (EWMA) method [Carter and Streilein 2012], Song et al. [2015]…”
MethodNon-Parametric Outliers Detection in Multiple Time Series A Case Study: Power Grid Data Analysis
Zhou, Y., Zou, H., Arghandeh, R., Gu, W., & Spanos, C. J. (2018) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v32i1.11632 · Cited by 26
A method for outlier detection based on cluster analysis and visual expert criteria
Lara, J. A., Lizcano, D., Rampérez, V., & Soriano, J. (2019) · Expert Systems · doi:10.1111/exsy.12473 · Cited by 26 · cited 3×
“2007], on bit representation clustering [Li et al. 2013], on bounding boxes [Sanchez and Bustos 2014], on clustering categorical data [Chau et al. 2018], and on the iSAX representation [Buu and Anh 2011]. Additionally, in Liu et al. [2009], not only is the HOT-SAX algorithm applied in a streaming context using a slidi…”
MethodInfluentialStochastic Online Anomaly Analysis for Streaming Time Series
Xu, Z., Kristian Kersting, B., & von Ritter, L. (2017) · International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2017/445 · Cited by 25
Online learning of Contextual Hidden Markov Models for temporal-spatial data analysis
Zhou, Y., Arghandeh, R., & Spanos, C. J. (2016) · Proceedings of the IEEE Conference on Decision & Control, including the Symposium on Adaptive Processes · doi:10.1109/cdc.2016.7799244 · Cited by 24 · cited 2×
“of past values. For example, the Contextual Hidden Markov Model (CHMM) incrementally captures both the temporal dependencies and the correlations between the variables in a multivariate time series [Zhou et al. 2016]. The temporal dependence is modeled by a basic HMM, and the correlation between the variables is inclu…”
MethodTime Series Discord Discovery Based on iSAX Symbolic Representation
Buu, H. T. Q., & Anh, D. T. (2011) · International Conference on Knowledge and Systems Engineering/Proceedings · doi:10.1109/kse.2011.11 · Cited by 21 · cited 3×
“[2009] Discord ✗ ✓ ✓ D Buu and Anh [2011] Discord ✗ ✓ ✓ D Chen and Cook [2011] Dissimilarity ✗ ✓ ✓ O/A Chen et al.”
BackgroundMethodNonparametric algorithm for identification of outliers in environmental data
Čampulová, M., Michálek, J., Mikuška, P., & Bokal, D. (2018) · Journal of Chemometrics · doi:10.1002/cem.2997 · Cited by 20
Efficient Detection of Discords for Time Series Stream
Liu, Y., Chen, X., Wang, F., & Yin, J. P. (2009) · Lecture notes in computer science · doi:10.1007/978-3-642-00672-2_62 · Cited by 17 · cited 1×
“Additionally, Liu et al. [2009] not only is the HOT-SAX algorithm applied in a streaming context using a sliding window but an incremental algorithm has also been suggested to simplify this computation.”
MethodECG Monitoring in Wearable Devices by Sparse Models
Carrera, D., Rossi, B., Zambon, D., Fragneto, P., & Boracchi, G. (2016) · Lecture notes in computer science · doi:10.1007/978-3-319-46131-1_21 · Cited by 14
Anomaly Detection in Streaming Time Series Based on Bounding Boxes
Sanchez, H., & Bustos, B. (2014) · Lecture notes in computer science · doi:10.1007/978-3-319-11988-5_19 · Cited by 13 · cited 2×
“[2015] Estimation ✗ ✓ ✓ ✓ O/A Song et al. [2015] Estimation ✗ ✓ ✓ ✓ Dirty Zhang et al.”
otsad: A package for online time-series anomaly detectors
Iturria, A., Carrasco, J., Charramendieta, S., Conde, Á., & Herrera, F. (2019) · Neurocomputing · doi:10.1016/j.neucom.2019.09.032 · Cited by 13 · cited 1×
“, Keogh et al. [2002]), equal-width binning (e.g., Chen and Cook [2011], Chen et al. [2013], and Rasheed and Alhajj [2014]), or SAX (e.”
BackgroundDiscord Discovery in Streaming Time Series based on an Improved HOT SAX Algorithm
Chau, P. M., Duc, B. M., & Anh, D. T. (2018) · doi:10.1145/3287921.3287929 · Cited by 9 · cited 2×
“…may be based on the Haar wavelet transformation and augmented trie [Bu et al. 2007; Fu et al. 2006], on bit representation clustering [Li et al. 2013], on bounding boxes [Sanchez and Bustos 2014], on clustering categorical data [Chau et al. 2018], and on the iSAX representation [Buu and Anh 2011].”
BackgroundMethodA Wearable Device for Online and Long-Term ECG Monitoring
Longoni, M., Carrera, D., Rossi, B., Fragneto, P., Pessione, M., & Boracchi, G. (2018) · International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2018/855 · Cited by 7 · cited 2×
“Subsequence outliers are those that cannot be accurately approximated by a linear combination of some subsequences of the dictionary (Ŝ), resulting in a large error in (7) using the Euclidean distance [Carrera et al. 2019; Longoni et al. 2018].”
MethodAutomatic outlier detection for time series
BasuSabyasachi, & MeckesheimerMartin (2007) · Knowledge and Information Systems · doi:10.5555/3225643.3225840 · Cited by 6
Applications of Outlier Analysis
Aggarwal, C. C. (2016) · Outlier Analysis · doi:10.1007/978-3-319-47578-3_13
Hawkins, D. M. (1980) · Identification of Outliers · doi:10.1007/978-94-015-3994-4_5
Research works printed but not linked: 6
- V. Chandola , A. Banerjee , and V. Kumar . 2007 . Outlier Detection: A Survey. Technical Report. Technical Report TR 07-017. Department of Computer Science and Engineering , University of Minnesota . V. Chandola, A. Banerjee, and V. Kumar. 2007. Outlier Detection: A Survey. Technical Report. Technical Report TR 07-017. Department of Computer Science and Engineering, University of Minnesota.
- M. Jones , D. Nikovski , M. Imamura , and T. Hirata . 2014. Anomaly detection in real-valued multidimensional time series . In Proceedings of the International Conference on Bigdata/Socialcom/Cybersecurity. Citeseer. M. Jones, D. Nikovski, M. Imamura, and T. Hirata. 2014. Anomaly detection in real-valued multidimensional time series. In Proceedings of the International Conference on Bigdata/Socialcom/Cybersecurity. Citeseer.
- 10.5555/3225652.3225900 · doi:10.5555/3225652.3225900
- 10.5555/2770961.2771087 · doi:10.5555/2770961.2771087
- P. Senin , J. Lin , X. Wang , T. Oates , S. Gandhi , A. P. Boedihardjo , C. Chen , and S. Frankenstein . 2015. Time series anomaly discovery with grammar-based compression . In Proceedings of 18th International Conference on Extending Database Technology (EDBT’15) . OpenProceedings.org, 481--492. P. Senin, J. Lin, X. Wang, T. Oates, S. Gandhi, A. P. Boedihardjo, C. Chen, and S. Frankenstein. 2015. Time series anomaly discovery with grammar-based compression. In Proceedings of 18th International Conference on Extending Database Technology (EDBT’15). OpenProceedings.org, 481--492.
- Z. Xu , K. Kersting , and L. von Ritter . 2016 . Adaptive streaming anomaly analysis . In Proceedings of NIPS Workshop on Artificial Intelligence for Data Science. Z. Xu, K. Kersting, and L. von Ritter. 2016. Adaptive streaming anomaly analysis. In Proceedings of NIPS Workshop on Artificial Intelligence for Data Science.
Cited by
Graph Neural Network-Based Anomaly Detection in Multivariate Time Series
Deng, A., & Hooi, B. (2021) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v35i5.16523 · Cited by 1,413
“A comprehensive summary is given in (Bl´azquez-Garc´ıa et al. 2020).”
BackgroundTransformers in Time Series: A Survey
Wen, Q., Zhou, T., Zhang, C., Chen, W., Ma, Z., Yan, J., & Sun, L. (2023) · Proceedings of the Thirty-Second International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2023/759 · Cited by 1,260
Deep Learning for Time Series Anomaly Detection: A Survey
Zamanzadeh Darban, Z., Webb, G. I., Pan, S., Aggarwal, C., & Salehi, M. (2025) · ACM Computing Surveys · doi:10.1145/3691338 · Cited by 613
“Although the field of anomaly detection has been explored in several literature surveys [140], [27], [20], [23] and some evaluation review papers exist [154], [102], there is only one survey on deep anomaly detection methods for time series data [41].”
Anomaly detection in time series: a comprehensive evaluation
Schmidl, S., Wenig, P., & Papenbrock, T. (2022) · Proceedings of the VLDB Endowment · doi:10.14778/3538598.3538602 · Cited by 590
Deep Learning for Anomaly Detection in Time-Series Data: Review, Analysis, and Guidelines
Choi, K., Yi, J., Park, C., & Yoon, S. (2021) · IEEE Access · doi:10.1109/access.2021.3107975 · Cited by 588
“To the best of our knowledge, previous reviews [1], [2], [9]–[14] on this subject matter do no more than simply categorize models according to their mechanisms and describe their characteristics.”
Background
Metadata
- Authors
- Blázquez-García, A., Conde, A., Mori, U., & Lozano, J. A.
- Affiliations
- Ikerlan Technology Research Centre, Basque Research and Technology Alliance (BRTA), Arrasate, Spain
- University of the Basque Country (ES)
- Venue
- ACM Computing Surveys
- Published
- 2021-04-17
- Volume
- 54
- Issue
- 3
- Pages
- 1-33
- ISSN
- 0360-0300 (ISSN-L)
- 1557-7341
- OpenAlex
- W3005893373
- Cited by
- 1,102
- References
- 105
- Citations by year
- 2026: 180 · 2025: 264 · 2024: 235 · 2023: 245 · 2022: 114 · 2021: 55
- FWCI
- 57.42
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- ACM Computing Surveys (publishedVersion, cc-by-nc-sa)
- BIRD (Basque Center for Applied Mathematics) (submittedVersion, cc-by-nc-sa)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- Ministerio de Ciencia, Innovación y Universidades (SEV-20 17-0718, BERC 2018-2021, PID2019)
- Eusko Jaurlaritza (KK-2019/00095, IT-1244-19, BERC 2018-2021, 2018‐2021, SEV-2017-0718, TIN2016-78365-R, BERC 2018‐2021 program)
- Basque Center for Applied Mathematics (SEV-2017-0718)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Anomaly Detection Techniques and Applications
- Physical Sciences › Computer Science › Signal Processing › Time Series Analysis and Forecasting
- Physical Sciences › Mathematics › Statistics and Probability › Advanced Statistical Methods and Models
- Keywords
- outlier detection, anomaly detection, time series, unsupervised learning, time series data mining
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.