计算机科学

计算机科学 ›› 2024, Vol. 51 ›› Issue (7): 108-115.doi: 10.11896/jsjkx.230400109

• 数据库&大数据&数据科学 • 上一篇    下一篇

缺失值场景下的多元时间序列异常检测算法

曾子辉1, 李超洋1,2, 廖清1,2   

  1. 1 哈尔滨工业大学(深圳)计算机科学与技术学院 广东 深圳 518055
    2 鹏城实验室 广东 深圳 518055
  • 收稿日期:2023-04-16 修回日期:2023-08-25 出版日期:2024-07-15 发布日期:2024-07-10
  • 通讯作者: 廖清(liaoqing@hit.edu.cn)
  • 作者简介:(20S151144@stu.hit.edu.cn)
  • 基金资助:
    国家自然科学基金面上项目(U1711261);广东省基础与应用基础研究重大项目(2019B030302002)

Multivariate Time Series Anomaly Detection Algorithm in Missing Value Scenario

ZENG Zihui1, LI Chaoyang1,2, LIAO Qing1,2   

  1. 1 School of Computer Science and Technology,Harbin Institute of Technology (Shenzhen),Shenzhen,Guangdong 518055,China
    2 Peng Cheng Laboratory,Shenzhen,Guangdong 518055,China
  • Received:2023-04-16 Revised:2023-08-25 Online:2024-07-15 Published:2024-07-10
  • About author:ZENG Zihui,born in 1997,postgra-duate.His main research interests include artificial intelligence and anomaly detection.
    LIAO Qing,born in 1988,Ph.D,professor,Ph.D supervisor.Her main research interests include artificial intelligence and data mining.
  • Supported by:
    National Natural Science Foundation of China(General Program)(U1711261) and Guangdong Basic and Applied Basic Research Foundation(Major Program)(2019B030302002).

PDF (PC)

457

摘要： 时间序列异常检测是工业界中一个重要的研究领域。当前的时间序列异常检测方法侧重于面向完整的时间序列数据进行异常检测,而没有考虑到包含工业场景中网络异常、传感器损坏等所导致的缺失值的时间序列异常检测任务。文中针对工业场景中更加常见的含缺失值的时间序列异常检测任务,提出了一种基于注意力重新表征的时间序列异常检测算法MMAD(Missing Multivariate Time Series Anomaly Detection)。具体来说,MMAD首先将包含缺失值的时间序列数据通过时间位置编码对时间序列中不同时间戳的空间关联进行建模,然后通过掩码注意力表征模块学习不同时间戳之间数据的关联关系并将其表征为一个高维的嵌入式编码矩阵,从而将包含缺失值的多元时间序列表示为不含缺失值的高维表征,最后引入条件标准化流对该表征进行重建,以重建概率作为异常评分,重建概率越小代表样本越异常。在3个经典时间序列数据集上进行实验,结果表明,相比其他基线方法,MMAD性能平均提升了11%,验证了MMAD在缺失值场景下进行多元时间序列异常检测的有效性。

关键词: 多元时间序列, 异常检测, 缺失值场景, 注意力机制, 神经网络

Abstract: Time series anomaly detection is an important research field in industry.Current methods of time series anomaly detection focus on anomaly detection for complete time series data,without considering the time series anomaly detection task containing missing values caused by network anomaly and sensor damage in industrial scenarios.In this paper,we propose an attention representation-based time series anomaly detection algorithm MMAD (missing multivariate time series anomaly detection) for the more common time series anomaly detection tasks with missing values in industrial scenarios.Specifically,MMAD first models the spatial correlation of different time stamps in time series by time position coding.Then,we build an attention representation module to learn the relationships between different time stamps and represent them as an embedded high-dimensional coding matrix,thereby representing the multivariate time series with missing values as a high-dimensional representation without missing values.Finally,we design the conditional normalized flow to reconstruct the representation and use the reconstruction probability as the anomaly score,the lower the probability of reconstruction,the more abnormal the sample.Experiments on three classical time series datasets show that,the average performance of MMAD is improved by 11% comparing with other baseline methods,which verifies the efficacy of MMAD to achieve multivariate time series anomaly detection with missing values.

Key words: Multivariate time series, Anomaly detection, Missing-value scenario, Attention mechanism, Neural network

中图分类号: 

  • TP389.1
[1]CHOI K,YI J,PARK C,et al.Deep Learning for Anomaly Detection in Time-series Data:Review,Analysis,and Guidelines[J].IEEE Access,2021,9:120043-120065.
[2]SUN Y,LI S H,CUI C,et al.Outlier Detection Method Basedon Gauss Calibration Function for Power User Data[J].Grid Technology,2018,42(5):1595-1606.
[3]PENG C,WANG L W,HU W L.Electromagnetic SpectrumAnomaly Detection Algorithm Based on Depth Feature Fusion [J].Journal of Electronics,2022,50(6):1359-1369.
[4]SUN C F,LU W M,DAI H D,et al.A Small Sample Data Augmentation Method Based on Time GAN and OCSVM for Multivariate Degenerated Equipment[J].Journal of Electronics,2022,50(11):2678-2687.
[5]ZHANG R B,ZUO Y C,ZHOU Z L,et al.Multimodal Generative Adversarial Networks Based Multivariate Time SeriesAnomaly Detection[J].Computer Science,2023,50(5):355-362.
[6]ZHANG J E,WU D,BOULET B.Time Series Anomaly Detection for Smart Grids:A Survey[C]//2021 IEEE Electrical Po-wer and Energy Conference(EPEC).IEEE,2021:125-130.
[7]PANG G,SHEN C,CAO L,et al.Deep Learning for Anomaly Detection:A Review[J].ACM Computing Surveys(CSUR),2021,54(2):1-38.
[8]WENING P,SCHMIDL S,PAPENBROCK T.TIMEEVAL:ABenchmarking Toolkit for Time Series Anomaly Detection Algorithms[J].Proceedings of the VLDB Endowment,2022,15(12):3678-3681.
[9]MUNIR M,SIDDIQUI S A,DENGEL A,et al.DEEPANT:A Deep Learning Approach for Unsupervised Anomaly Detection in Time Series[J].IEEE Access,2018,7:1991-2005.
[10]HUNDMAN K,CONSTANTINOU V,LAPORTE C,et al.Detecting Spacecraft Anomalies Using LSTMs and Nonparametric Dynamic Thresholding[C]//Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining.2018:387-395.
[11]SU Y,ZHAO Y,NIU C,et al.Robust Anomaly Detection for Multivariate Time Series Through Stochastic Recurrent Neural Network[C]//Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining.2019:2828-2837.
[12]CHEN Z,CHEN D,ZHANG X,et al.Learning Graph Structures with Transformer for Multivariate Time-Series Anomaly Detection in IOT[J].IEEE Internet of Things Journal,2021,9(12):9179-9189.
[13]VASWANI A,SHAZEER N,PARMAR N,et al.Attention isAll You Need[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems.2017:6000-6010.
[14]DENG A,HOOI B.Graph NeuralNetwork-Based Anomaly Detection in Multivariate Time Series[C]//Proceedings of the AAAI Conference on Artificial Intelligence.2021:4027-4035.
[15]ZONG B,SONG Q,MIN M R,et al.Deep Autoencoding Gaus-sian Mixture Model for Unsupervised Anomaly Detection[C]//International Conference on Learning Representations.2018.
[16]DAI E Y,CHEN J.Graph-Augmented Normalizing Flows for Anomaly Detection of Multiple Time Series[C]//International Conference on Learning Representations.2022.
[17]AUDIBERT J,MICHIARDI P,GUYARD F,et al.USAD:Unsupervised Anomaly Detection on Multivariate Time Series[C]//Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining.2020:3395-3404.
[18]CHEN X,DENG L,HUANG F,et al.DAEMON:Unsupervised Anomaly Detection and Interpretation for Multivariate Time Series[C]//2021 IEEE 37th International Conference on Data Engineering (ICDE).IEEE,2021:2225-2230.
[19]LI D,CHEN D,JIN B,et al.MAD-GAN:Multivariate Anomaly Detection for Time Series Data with Generative Adversarial Networks[C]//Artificial Neural Networks and Machine Lear-ning－ICANN.2019:703-716.
[20]CHEN S W,LI J,XUAN J X,et al.LSTM-GAN:Unsupervised Anomaly Detection for Time Series Fusion of GAN and Bi-LSTM[J].Journal of Chinese Computer Systems,2024,45(1):123-131.
[21]REZENDE D,MOHAMED S.Variational Inference with Nor-malizing Flows[C]//International Conference on Machine Learning.PMLR,2015:1530-1538.
[22]GOH J,ADEPU S,JUNEJO K N,et al.A Dataset to Support Research in The Design of Secure Water Treatment Systems[C]//International Conference on Critical Information Infrastructures Security.Springer,2016:88-99.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发