{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,2]],"date-time":"2026-06-02T14:36:51Z","timestamp":1780411011457,"version":"3.54.1"},"reference-count":40,"publisher":"MDPI AG","issue":"11","license":[{"start":{"date-parts":[[2022,11,5]],"date-time":"2022-11-05T00:00:00Z","timestamp":1667606400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["52231014"],"award-info":[{"award-number":["52231014"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Entropy"],"abstract":"<jats:p>To ensure the normal operation of the system, the enterprise\u2019s operations engineer will monitor the system through the KPI (key performance indicator). For example, web page visits, server memory utilization, etc. KPI anomaly detection is a core technology, which is of great significance for rapid fault detection and repair. This paper proposes a novel dual-stage attention-based LSTM-VAE (DA-LSTM-VAE) model for KPI anomaly detection. Firstly, in order to capture time correlation in KPI data, long\u2013short-term memory (LSTM) units are used to replace traditional neurons in the variational autoencoder (VAE). Then, in order to improve the effect of KPI anomaly detection, an attention mechanism is introduced into the input stage of the encoder and decoder, respectively. During the input stage of the encoder, a time attention mechanism is adopted to assign different weights to different time points, which can adaptively select important input sequences to avoid the influence of noise in the data. During the input stage of the decoder, a feature attention mechanism is adopted to adaptively select important latent variable representations, which can capture the long-term dependence of time series better. In addition, this paper proposes an adaptive threshold method based on anomaly scores measured by reconstruction probability, which can minimize false positives and false negatives and avoid adjustment of the threshold manually. Experimental results in a public dataset show that the proposed method in this paper outperforms other baseline methods.<\/jats:p>","DOI":"10.3390\/e24111613","type":"journal-article","created":{"date-parts":[[2022,11,8]],"date-time":"2022-11-08T11:46:43Z","timestamp":1667908003000},"page":"1613","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":10,"title":["DA-LSTM-VAE: Dual-Stage Attention-Based LSTM-VAE for KPI Anomaly Detection"],"prefix":"10.3390","volume":"24","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-4563-4475","authenticated-orcid":false,"given":"Yun","family":"Zhao","sequence":"first","affiliation":[{"name":"School of Information Science and Technology, Dalian Maritime University, Dalian 116026, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiuguo","family":"Zhang","sequence":"additional","affiliation":[{"name":"School of Information Science and Technology, Dalian Maritime University, Dalian 116026, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zijing","family":"Shang","sequence":"additional","affiliation":[{"name":"School of Information Science and Technology, Dalian Maritime University, Dalian 116026, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zhiying","family":"Cao","sequence":"additional","affiliation":[{"name":"School of Information Science and Technology, Dalian Maritime University, Dalian 116026, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2022,11,5]]},"reference":[{"key":"ref_1","first-page":"68","article-title":"Intelligent operation and maintenance based on machine learning","volume":"13","author":"Pei","year":"2017","journal-title":"Commun. CCF"},{"key":"ref_2","doi-asserted-by":"crossref","unstructured":"He, S., Yang, B., and Qiao, Q. (2021, January 23\u201325). Overview of Key Performance Indicator Anomaly Detection. Proceedings of the IEEE Region 10 Symposium (TENSYMP), Jeju, Republic of Korea.","DOI":"10.1109\/TENSYMP52854.2021.9550989"},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"032401","DOI":"10.1115\/1.4037963","article-title":"Optimization of statistical methodologies for anomaly detection in gas turbine dynamic time series","volume":"140","author":"Ceschini","year":"2018","journal-title":"J. Eng. Gas Turbines Power"},{"key":"ref_4","doi-asserted-by":"crossref","unstructured":"Reis, B.Y., and Mandl, K.D. (2003). Time series modeling for syndromic surveillance. BMC Med. Inform. Decis. Mak., 3.","DOI":"10.1186\/1472-6947-3-2"},{"key":"ref_5","unstructured":"Malhotra, P., Vig, L., Shroff, G., and Agarwal, P. (2015, January 22\u201324). Long Short Term Memory Networks for Anomaly Detection in Time Series. Proceedings of the 23rd European Symposium on Artificial Neural Networks (ESANN), Bruges, Belgium."},{"key":"ref_6","doi-asserted-by":"crossref","unstructured":"Ghanbari, M., Kinsner, W., and Ferens, K. (2016, January 12\u201314). Anomaly detection in a smart grid using wavelet transform, variance fractal dimension and an artificial neural network. Proceedings of the IEEE Electrical Power and Energy Conference (EPEC), Ottawa, ON, Canada.","DOI":"10.1109\/EPEC.2016.7771715"},{"key":"ref_7","doi-asserted-by":"crossref","unstructured":"Laptev, N., Amizadeh, S., and Flint, I. (2015, January 10\u201313). Generic and Scalable Framework for Automated Time-series Anomaly Detection. Proceedings of the 21st ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD), Sydney, Australia.","DOI":"10.1145\/2783258.2788611"},{"key":"ref_8","doi-asserted-by":"crossref","unstructured":"Liu, D., Zhao, Y., Xu, H., Sun, Y., Pei, D., Luo, J., Jing, X., and Feng, M. (2015, January 28\u201330). Opprentice: Towards practical and automatic anomaly detection through machine learning. Proceedings of the ACM Measurement Conference, Tokyo, Japan.","DOI":"10.1145\/2815675.2815679"},{"key":"ref_9","doi-asserted-by":"crossref","unstructured":"Shi, J., He, G., and Liu, X. (2018, January 22\u201324). Anomaly Detection for Key Performance Indicators through Machine Learning. Proceedings of the 6th IEEE International Conference on Network Infrastructure and Digital Content (NIDC), Guiyang, China.","DOI":"10.1109\/ICNIDC.2018.8525714"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"121","DOI":"10.1016\/j.patcog.2016.03.028","article-title":"High-dimensional and large-scale anomaly detection using a linear one-class SVM with deep learning","volume":"58","author":"Erfani","year":"2016","journal-title":"Pattern Recognit."},{"key":"ref_11","unstructured":"Yang, X., Latecki, L.J., and Pokrajac, D. (May, January 30). Outlier Detection with Globally Optimal Exemplar-Based GMM. Proceedings of the Siam International Conference on Data Mining (SDM), Sparks, NV, USA."},{"key":"ref_12","first-page":"1","article-title":"Variational autoencoder based anomaly detection using reconstruction probability","volume":"2","author":"An","year":"2015","journal-title":"Spec. Lect. IE"},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Kawachi, Y., Koizumi, Y., and Harada, N. (2018, January 15\u201320). Complementary Set Variational Autoencoder for Supervised Anomaly Detection. Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Calgary, AB, Canada.","DOI":"10.1109\/ICASSP.2018.8462181"},{"key":"ref_14","unstructured":"Xu, H., Feng, Y., Chen, J., Wang, Z., Qiao, H., and Chen, W. (2018, January 23\u201327). Unsupervised Anomaly Detection via Variational Auto-Encoder for Seasonal KPIs in Web Applications. Proceedings of the 27th World Wide Web (WWW) Conference, Lyon, France."},{"key":"ref_15","doi-asserted-by":"crossref","unstructured":"Li, Z., Chen, W., and Pei, D. (2018, January 17\u201319). Robust and Unsupervised KPI Anomaly Detection Based on Conditional Variational Autoencoder. Proceedings of the 37th International Performance Computing and Communications Conference (IPCCC), Orlando, FL, USA.","DOI":"10.1109\/PCCC.2018.8710885"},{"key":"ref_16","unstructured":"Kingma, D.P., and Welling, M. (2014, January 14\u201316). Auto-encoding variational bayes. Proceedings of the 2nd International Conference on Learning Representations (ICLR), Banff, AB, Canada."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"2746","DOI":"10.1109\/JSAC.2022.3191341","article-title":"Situation-Aware Multivariate Time Series Anomaly Detection Through Active Learning and Contrast VAE-Based Models in Large Distributed Systems","volume":"40","author":"Li","year":"2022","journal-title":"IEEE J. Sel. Areas Commun."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"2440","DOI":"10.1109\/JSAC.2022.3180785","article-title":"Efficient KPI Anomaly Detection Through Transfer Learning for Large-Scale Web Services","volume":"40","author":"Zhang","year":"2022","journal-title":"IEEE J. Sel. Areas Commun."},{"key":"ref_19","unstructured":"Sutskever, I., Vinyals, O., and Le, Q.V. (2014, January 8\u201313). Sequence to Sequence Learning with Neural Networks. Proceedings of the 27th International Conference on Neural Information Processing Systems, Montreal, QC, Canada."},{"key":"ref_20","unstructured":"Malhotra, P., Ramakrishnan, A., Anand, G., Vig, L., and Shroff, G. (2016, January 24). LSTM-based Encoder-Decoder for Multi-sensor Anomaly Detection. Presented at ICML 2016 Anomaly Detection Workshop, New York, NY, USA."},{"key":"ref_21","first-page":"1544","article-title":"A Multimodal Anomaly Detector for Robot-Assisted Feeding Using an LSTM-based Variational Autoencoder","volume":"3","author":"Daehyung","year":"2017","journal-title":"IEEE Robot. Autom. Lett."},{"key":"ref_22","doi-asserted-by":"crossref","unstructured":"Lin, S., Clark, R., Birke, R., Schonborn, S., and Roberts, S. (2020, January 4\u20138). Anomaly Detection for Time Series Using VAE-LSTM Hybrid Model. Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Barcelona, Spain.","DOI":"10.1109\/ICASSP40776.2020.9053558"},{"key":"ref_23","doi-asserted-by":"crossref","unstructured":"Su, Y., Zhao, Y., Niu, C., Liu, R., and Pei, D. (2019, January 4\u20138). Robust Anomaly Detection for Multivariate Time Series through Stochastic Recurrent Neural Network. Proceedings of the 25th International Conference on Knowledge Discovery Data Mining (KDD), Anchorage, AK, USA.","DOI":"10.1145\/3292500.3330672"},{"key":"ref_24","doi-asserted-by":"crossref","unstructured":"Niu, Z., Yu, K., and Wu, X. (2020). LSTM-Based VAE-GAN for Time-Series Anomaly Detection. Sensors, 20.","DOI":"10.3390\/s20133738"},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"130","DOI":"10.1016\/j.neucom.2021.03.062","article-title":"A joint model for anomaly detection and trend prediction on it operation series","volume":"448","author":"Chen","year":"2021","journal-title":"Neurocomputing"},{"key":"ref_26","doi-asserted-by":"crossref","unstructured":"Chen, N., Tu, H., Duan, X., Hu, L., and Guo, C. (2022). Semisupervised anomaly detection of multivariate time series based on a variational autoencoder. Appl. Intell.","DOI":"10.1007\/s10489-022-03829-1"},{"key":"ref_27","unstructured":"Bahdanau, D., Cho, K., and Bengio, Y. (2015, January 7\u20139). Neural Machine Translation by Jointly Learning to Align and Translate. Proceedings of the 3rd International Conference on Learning Representations (ICLR), San Diego, CA, USA."},{"key":"ref_28","unstructured":"Vaswani, A., Shazeer, N., and Parmar, N. (2017, January 4\u20139). Attention Is All You Need. Proceedings of the 31st International Conference on Neural Information Processing Systems, Long Beach, CA, USA."},{"key":"ref_29","unstructured":"Zhang, C., Song, D., and Chen, Y. (February, January 27). A Deep Neural Network for Unsupervised Anomaly Detection and Diagnosis in Multivariate Time Series Data. Proceedings of the 33rd AAAI Conference on Artificial Intelligence\/31st Innovative Applications of Artificial Intelligence Conference\/9th AAAI Symposium on Educational Advances in Artificial Intelligence, Hawaii, HI, USA."},{"key":"ref_30","doi-asserted-by":"crossref","unstructured":"Pereira, J., and Silveira, M. (2018, January 17\u201320). Unsupervised Anomaly Detection in Energy Time Series Data Using Variational Recurrent Autoencoders with Attention. Proceedings of the 17th IEEE International Conference on Machine Learning and Applications (ICMLA), Orlando, FL, USA.","DOI":"10.1109\/ICMLA.2018.00207"},{"key":"ref_31","unstructured":"Yao, Q., Song, D., and Chen, H. (2017, January 19\u201325). A Dual-Stage Attention-Based Recurrent Neural Network for Time Series Prediction. Proceedings of the 26th International Joint Conference on Artificial Intelligence (IJCAI), Melbourne, Australia."},{"key":"ref_32","doi-asserted-by":"crossref","first-page":"558","DOI":"10.1080\/01621459.1995.10476548","article-title":"Using Markov Chain Monte Carlo","volume":"90","author":"Rosenthal","year":"1995","journal-title":"J. Am. Stat. Assoc."},{"key":"ref_33","doi-asserted-by":"crossref","first-page":"759","DOI":"10.1037\/a0019471","article-title":"A Dual-Stage Two-Phase Model of Selective Attention","volume":"117","author":"Steinhauser","year":"2010","journal-title":"Psychol. Rev."},{"key":"ref_34","doi-asserted-by":"crossref","first-page":"113082","DOI":"10.1016\/j.eswa.2019.113082","article-title":"DSTP-RNN: A dual-stage two-phase attention-based recurrent neural network for long-term and multivariate time series prediction","volume":"143","author":"Liu","year":"2020","journal-title":"Expert Syst. Appl."},{"key":"ref_35","doi-asserted-by":"crossref","first-page":"1735","DOI":"10.1162\/neco.1997.9.8.1735","article-title":"Long Short-Term Memory","volume":"9","author":"Hochreiter","year":"1997","journal-title":"Neural Comput."},{"key":"ref_36","doi-asserted-by":"crossref","first-page":"504","DOI":"10.1126\/science.1127647","article-title":"Reducing the Dimensionality of data with neural networks","volume":"313","author":"Hinton","year":"2006","journal-title":"Science"},{"key":"ref_37","doi-asserted-by":"crossref","unstructured":"Bowman, S.R., Vilnis, L., and Vinyals, O. (2016, January 11\u201312). Generating Sentences from a Continuous Space. Proceedings of the 20th SIGNLL Conference on Computational Natural Language Learning, Berlin, Germany.","DOI":"10.18653\/v1\/K16-1002"},{"key":"ref_38","doi-asserted-by":"crossref","unstructured":"Deng, M., and Wu, B. (2020, January 12\u201314). Self-adaptive Threshold Traffic Anomaly Detection Based on \u03c6-Entropy and the Improved EWMA Model. Proceedings of the 4th Information Technology, Networking, Electronic and Automation Control Conference (ITNEC), ELECTR NETWORK, Chongqing, China.","DOI":"10.1109\/ITNEC48623.2020.9084673"},{"key":"ref_39","first-page":"2812","article-title":"Principal Component Analysis","volume":"6","author":"Bro","year":"2014","journal-title":"J. Mark. Res."},{"key":"ref_40","first-page":"2579","article-title":"Visualizing Data using t-SNE","volume":"9","author":"Laurens","year":"2008","journal-title":"J. Mach. Learn. Res."}],"container-title":["Entropy"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1099-4300\/24\/11\/1613\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,11]],"date-time":"2025-10-11T01:11:09Z","timestamp":1760145069000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1099-4300\/24\/11\/1613"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,11,5]]},"references-count":40,"journal-issue":{"issue":"11","published-online":{"date-parts":[[2022,11]]}},"alternative-id":["e24111613"],"URL":"https:\/\/doi.org\/10.3390\/e24111613","relation":{},"ISSN":["1099-4300"],"issn-type":[{"value":"1099-4300","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,11,5]]}}}