DeepAnT: A Deep Learning Approach for Unsupervised Anomaly Detection in Time Series

By Mohsin Munir, Shoaib Ahmed Siddiqui, Andreas R. Dengel, Sheraz Ahmed. IEEE Access, 2019

618 citations recorded by the source corpus.

Traditional distance and density-based anomaly detection techniques are unable to detect periodic and seasonality related point anomalies which occur commonly in streaming data, leaving a big gap in time series anomaly detection in the current era of the IoT. To address this problem, we present a novel deep learning-based anomaly detection approach (DeepAnT) for time series data, which is equally applicable to the non-streaming cases. DeepAnT is capable of detecting a wide range of anomalies, i.e., point anomalies, contextual anomalies, and discords in time series data. In contrast to the anomaly detection methods where anomalies are learned, DeepAnT uses unlabeled data to capture and learn the data distribution that is used to forecast the normal behavior of a time series. DeepAnT consists of two modules: time series predictor and anomaly detector. The time series predictor module uses deep convolutional neural network (CNN) to predict the next time stamp on the defined horizon. This module takes a window of time series (used as a context) and attempts to predict the next time stamp. The predicted value is then passed to the anomaly detector module, which is responsible for tagging the corresponding time stamp as normal or abnormal. DeepAnT can be trained even without removing the anomalies from the given data set. Generally, in deep learning-based approaches, a lot of data are required to train a model. Whereas in DeepAnT, a model can be trained on relatively small data set while achieving good generalization capabilities due to the effective parameter sharing of the CNN. As the anomaly detection in DeepAnT is unsupervised, it does not rely on anomaly labels at the time of model generation. Therefore, this approach can be directly applied to real-life scenarios where it is practically impossible to label a big stream of data coming from heterogeneous sensors comprising of both normal as well as anomalous points. We have performed a detailed evaluation of 15 algorithms on 10 anomaly detection benchmarks, which contain a total of 433 real and synthetic time series. Experiments show that DeepAnT outperforms the state-of-the-art anomaly detection methods in most of the cases, while performing on par with others.

Subfields: Artificial Intelligence, Computer Networks and Communications, Signal Processing

View Paper: https://www.semanticscholar.org/paper/f37bc75aa1833e1330c39c4f04b131baca08d67b

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Networks and Communications

In Signal Processing

Citations in the corpus

A Review on Outlier/Anomaly Detection in Time Series Data (2020, methodology): “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.”
Deep Learning for Time Series Anomaly Detection: A Survey (2022, neutral): “2) DeepAnt [133] T Un W Point + Subseq TCN-ms [77] T Semi W Subseq”
Learning Graph Structures With Transformer for Multivariate Time-Series Anomaly Detection in IoT (2021, methodology): “DeepAnt [31] was an unsupervised approach using convolutional neural network (CNN) to forecast future time series values and adopted Euclidean distance to measure the discrepancy for anomaly detection.”
Anomaly Detection in Time Series: A Comprehensive Evaluation (2022, methodology): “The representatives of this category that we consider in our evaluation are: AD-LTI [148], ARIMA [65], DeepAnT [94], DeepNAP [72], HealthESN [32], LSTMAD [89], MedianMethod [10], MTAD-GAT [161], NumentaHTM [3], NoveltySVR [86], OceanWNN [143], RBForest [165], RForest [21], SARIMA [52], Telemanom [64”
Graph Neural Network-Based Anomaly Detection in Multivariate Time Series (2021, methodology): “These techniques, such as Convolutional Neural Network (CNN) based models (Munir et al. 2018), Long Short Term Memory (LSTM) (Filonov, Lavrentyev, and Vorontsov 2016; Hund-man et al. 2018a; Park, Hoshi, and Kemp 2018) and Generative Adversarial Networks (GAN) models (Zhou et al. 2019; Li et al.…”
Pyraformer: Low-Complexity Pyramidal Attention for Long-Range Time Series Modeling and Forecasting (2022, background): “As mentioned in the previous section, CNN and RNN enjoy a low time and space complexity (i.e., O ( L ) ), but entail a path of O ( L ) to describe long-range dependence.”
Deep Anomaly Detection for Time-Series Data in Industrial IoT: A Communication-Efficient On-Device Federated Learning Approach (2020, methodology): “[15] designed a novel DAD approach, called DeepAnT, to achieve anomaly detection by utilizing the deep convolutional neural network (CNN) to predict anomaly value.”
Deep Learning for Anomaly Detection in Time-Series Data: Review, Analysis, and Guidelines (2021, positive): “Mean-while, CNN-based approaches [70], [71] have been veri-ﬁed to be effective on several datasets from global cloud enterprises.”
An Evaluation of Anomaly Detection and Diagnosis in Multivariate Time Series (2021, methodology): “Normalized errors: Errors, Ert can be used directly as the anomaly score [45].”