Online Few-Shot Time Series Classification for Aftershock Detection
Pages 5707 - 5716
Abstract
Seismic monitoring systems sift through seismograms in real-time, searching for target events, such as underground explosions. In this monitoring system, a burst of aftershocks (minor earthquakes occur after a major earthquake over days or even years) can be a source of confounding signals. Such a burst of aftershock signals can overload the human analysts of the monitoring system. To alleviate this burden at the onset of a sequence of events (e.g., aftershocks), a human analyst can label the first few of these events and start an online classifier to filter out subsequent aftershock events. We propose an online few-shot classification model FewSig for time series data for the above use case. The framework of FewSig consists of a selective model to identify the high-confidence positive events which are used for updating the models and a general classifier to label the remaining events. Our specific technique uses a %two-level decision tree selective model based on sliding DTW distance and a general classifier model based on distance metric learning with Neighborhood Component Analysis (NCA). The algorithm demonstrates surprising robustness when tested on univariate datasets from the UEA/UCR archive. Furthermore, we show two real-world earthquake events where the FewSig reduces the human effort in monitoring applications by filtering out the aftershock events.
Supplementary Material
When a large earthquake occurs, it unleashes widespread devastation and disrupts our crucial seismic monitoring system. The current system relies heavily on human analysts to verify algorithm-generated results. However, in the aftermath of a major earthquake, the numerous aftershocks can overwhelm analysts, making it challenging for them to keep pace with the workload. This vulnerability jeopardizes the ability to detect critical signals, including potential nuclear tests, within the crucial timeframe. Fortunately, Our team has developed an innovative online few-shot time series classification model named Fewsig. It learns the signature from the initial five aftershocks, and can accurately identify the subsequent aftershocks of the current earthquake in real-time. Analysts in the monitoring system can use Fewsig to filter out unimportant aftershocks and focus on more critical signals.
- Download
- 146.63 MB
References
[1]
1999. IDC Documentation, Processing of Seismic, Hydroacoustic, and Infrasonic Data (1999), (IDC/OPS/MA/001/Rev.1).
[2]
Rex V Allen. 1978. Automatic earthquake recognition and timing from single traces. Bulletin of the seismological society of America, Vol. 68, 5 (1978), 1521--1532.
[3]
William Vickers Anthony Bagnall, Jason Lines and Eamonn Keogh. 2023. The UEA & UCR Time Series Classification Repository. www.timeseriesclassification.com Retrieved Jan 19, 2023 from
[4]
Yanping Chen, Bing Hu, Eamonn Keogh, and Gustavo E.A.P.A Batista. 2013. DTW-D: Time Series Semi-Supervised Learning from a Single Example. In Proceedings of the KDD 2013 (Chicago, Illinois, USA) (KDD '13). Association for Computing Machinery, New York, NY, USA, 383--391. https://doi.org/10.1145/2487575.2487633
[5]
Andrea Conley, Brendan Donohoe, and Benjamin Greene. 2021. Aftershock Identification Using a Paired Neural Network Applied to Constructed Data. (6 2021). https://doi.org/10.2172/1821802
[6]
Haoyi Fan, Fengbin Zhang, Ruidong Wang, Xunhua Huang, and Zuoyong Li. 2021. Semi-Supervised Time Series Classification by Temporal Relation Prediction. IEEE International Conference on Acoustics, Speech and Signal Processing, 3545--3549. https://doi.org/10.1109/ICASSP39728.2021.9413883
[7]
Chelsea Finn, Pieter Abbeel, and Sergey Levine. 2017. Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks. In Proceedings of the 34th International Conference on Machine Learning (Proceedings of Machine Learning Research, Vol. 70), Doina Precup and Yee Whye Teh (Eds.). PMLR, 1126--1135. https://proceedings.mlr.press/v70/finn17a.html
[8]
Rafael Giusti, Diego F. Silva, and Gustavo E. A. P. A. Batista. 2016. Improved Time Series Classification with Representation Diversity and SVM. In Proceedings of the ICMLA 2016. 1--6. https://doi.org/10.1109/ICMLA.2016.0010
[9]
Jacob Goldberger, Sam Roweis, Geoff Hinton, and Ruslan Salakhutdinov. 2004. Neighbourhood Components Analysis. In International Conference on Neural Information Processing Systems (Vancouver, British Columbia, Canada) (NIPS'04). MIT Press, Cambridge, MA, USA, 513--520.
[10]
Yunhui Guo, Noel C. Codella, Leonid Karlinsky, James V. Codella, John R. Smith, Kate Saenko, Tajana Rosing, and Rogerio Feris. 2020. A Broader Study of Cross-Domain Few-Shot Learning. In Computer Vision - ECCV 2020, Andrea Vedaldi, Horst Bischof, Thomas Brox, and Jan-Michael Frahm (Eds.). Springer International Publishing, Cham, 124--141.
[11]
Brijnesh Jain and Stephan Spiegel. 2016. Dimension Reduction in Dissimilarity Spaces for Time Series Classification. In Advanced Analysis and Learning on Temporal Data. Springer, Cham, 31--46.
[12]
Shayan Jawed, Josif Grabocka, and Lars Schmidt-Thieme. 2020. Self-supervised learning for semi-supervised time series classification. In Pacific-Asia Conference on KDD. Springer, 499--511.
[13]
Rohit J. Kate. 2016. Using dynamic time warping distances as features for improved time series classification. Data Mining and Knowledge Discovery, Vol. 30 (3 2016), 283--312. Issue 2. https://doi.org/10.1007/s10618-015-0418-x
[14]
Tsung-Yi Lin, Priya Goyal, Ross Girshick, Kaiming He, and Piotr Dollar. 2020. Focal Loss for Dense Object Detection. IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 42 (2 2020), 318--327. Issue 2. https://doi.org/10.1109/TPAMI.2018.2858826
[15]
Anthony Lomax, Claudio Satriano, and Maurizio Vassallo. 2012. Automatic picker developments and optimization. Seismological Research Letters, Vol. 83, 3 (2012), 531--540.
[16]
Kristóf Marussy and Krisztian Buza. 2013. SUCCESS: A New Approach for Semi-supervised Classification of Time-Series. In Artificial Intelligence and Soft Computing. Springer Berlin Heidelberg, Berlin, Heidelberg, 437--447.
[17]
D.E. McNamara, W.L. Yeck, W.D. Barnhart, V. Schulte-Pelkum, E. Bergman, L.B. Adhikari, A. Dixit, S.E. Hough, H.M. Benz, and P.S. Earle. 2017. Source modeling of the 2015 Mw 7.8 Nepal (Gorkha) earthquake sequence. Tectonophysics, Vol. 714--715 (2017), 21--30. https://doi.org/10.1016/j.tecto.2016.08.004 Special Issue on the 25 April 2015 Mw 7.8 Gorkha (Nepal) Earthquake.
[18]
MexicoSSN. 2016. Mexico Earthquake catalog. http://www2.ssn.unam.mx:8080/catalogo Retrieved Jan 19, 2023 from
[19]
Abdullah Mueen. 2016a. Extracting Optimal Performance from Dynamic Time Warping. https://www.cs.unm.edu/ mueen/DTW.pdf Retrieved Jan 19, 2023 from
[20]
Abdullah Mueen. 2016b. Similarity Search on Time Series Data: Past, Present and Future. https://www.cs.unm.edu/ mueen/Tutorial/CIKM2016Tutorial.pdf Retrieved Jan 19, 2023 from
[21]
Abdullah Mueen, Eamonn Keogh, and Neal Young. 2011. Logical-shapelets: an expressive primitive for time series classification. Proceedings of the KDD 2011 (2011), 1154--1162. https://doi.org/10.1145/2020408.2020587
[22]
Jyoti Narwariya, Pankaj Malhotra, Lovekesh Vig, Gautam Shroff, and TV Vishnu. 2020. Meta-learning for few-shot time series classification. In ACM IKDD CoDS. 28--36.
[23]
Minh Nhut Nguyen, Xiao-Li Li, and See-Kiong Ng. 2011. Positive Unlabeled Learning for Time Series Classification. In International Joint Conference on Artificial Intelligence (Barcelona, Catalonia, Spain) (IJCAI'11). AAAI Press, 1421--1426.
[24]
Alex Nichol, Joshua Achiam, and John Schulman. 2018. On First-Order Meta-Learning Algorithms. (3 2018).
[25]
Brian Pope. 2022. Air Force Technological Advancement Center. Personal Communication. March 9, 2022 at 12:33 PM.
[26]
Thanawin Rakthanmanon, Bilson Campana, Abdullah Mueen, Gustavo Batista, Brandon Westover, Qiang Zhu, Jesin Zakaria, and Eamonn Keogh. 2012. Searching and mining trillions of time series subsequences under dynamic time warping. In Proceedings of the KDD 2012. 262--270. https://doi.org/10.1145/2339530.2339576
[27]
Christos Saragiotis. 2021. Efforts toward automatic aftershock sequences processing at the International Data Centre. In EGU General Assembly Conference Abstracts (EGU General Assembly Conference Abstracts). Article EGU21--5062, EGU21--5062 pages. https://doi.org/10.5194/egusphere-egu21-5062
[28]
Jake Snell, Kevin Swersky, and Richard Zemel. 2017. Prototypical Networks for Few-Shot Learning. In Proceedings of the 31st International Conference on Neural Information Processing Systems (Long Beach, California, USA) (NIPS'17). Curran Associates Inc., Red Hook, NY, USA, 4080--4090.
[29]
C. A. R. Sousa, V. M. A. Souza, and Gustavo Batista. 2014. Time series transductive classification on imbalanced data sets: an experimental study. In International Conference on Pattern Recognition. IEEE, 3780--3785.
[30]
Vinicius M. A. Souza, Rafael G. Rossi, Gustavo Batista, and Solange O Rezende. 2017. Unsupervised active learning techniques for labeling training sets: an experimental evaluation on sequential data. Intelligent Data Analysis, Vol. 21, 5 (2017), 1061--1095.
[31]
Flood Sung, Yongxin Yang, Li Zhang, Tao Xiang, Philip HS Torr, and Timothy M Hospedales. 2018. Learning to compare: Relation network for few-shot learning. In IEEE conference on computer vision and pattern recognition. 1199--1208.
[32]
Haishuai Wang, Qin Zhang, Jia Wu, Shirui Pan, and Yixin Chen. 2019. Time series feature learning with labeled and unlabeled data. Pattern Recognition, Vol. 89 (2019), 55--66. https://doi.org/10.1016/j.patcog.2018.12.026
[33]
Yaqing Wang, Quanming Yao, James T Kwok, and Lionel M Ni. 2020. Generalizing from a few examples: A survey on few-shot learning. ACM computing surveys, Vol. 53, 3 (2020), 1--34.
[34]
Li Wei and Eamonn Keogh. 2006. Semi-supervised time series classification. In Proceedings of the KDD 2006. ACM Press, New York, New York, USA, 748. https://doi.org/10.1145/1150402.1150498
[35]
Liang Xi, Zichao Yun, Han Liu, Ruidong Wang, Xunhua Huang, and Haoyi Fan. 2022. Semi-supervised Time Series Classification Model with Self-supervised Learning. Engineering Applications of Artificial Intelligence, Vol. 116 (2022), 105331. https://doi.org/10.1016/j.engappai.2022.105331
[36]
Zhao Xu and Koichi Funaya. 2015. Time series analysis with graph-based semi-supervised learning. In IEEE international conference on data science and advanced analytics. IEEE, 1--6.
[37]
Qi Zhang, Tong Xu, Hengshu Zhu, Lifu Zhang, Hui Xiong, Enhong Chen, and Qi Liu. 2019. Aftershock Detection with Multi-scale Description Based Neural Network. In 2019 IEEE International Conference on Data Mining (ICDM). 886--895. https://doi.org/10.1109/ICDM.2019.00099
[38]
Sheng Zhong. 2022. Support Website. https://sites.google.com/view/onlinefewshot/home Retrieved Jan 19, 2023 from https://sites.google.com/view/onlinefewshot/home
Index Terms
- Online Few-Shot Time Series Classification for Aftershock Detection
Recommendations
Near-Real-Time Seismic Human Fatality Information Retrieval from Social Media with Few-Shot Large-Language Models
SenSys '22: Proceedings of the 20th ACM Conference on Embedded Networked Sensor SystemsReal-time disaster-induced human fatality information is critical for rapid and accurate disaster impact and loss estimation and effective emergency response. Systems like PAGER incorporate online reported death tolls and loss projection models trained ...
Self-labeling techniques for semi-supervised time series classification: an empirical study
An increasing amount of unlabeled time series data available render the semi-supervised paradigm a suitable approach to tackle classification problems with a reduced quantity of labeled data. Self-labeled techniques stand out from semi-supervised ...
Time series classification, augmentation and artificial-intelligence-enabled software for emergency response in freight transportation fires
AbstractIn responding to freight transportation fire incidents, first responders refer to the terials labeled on the freights and the Emergency Response Guidebook (ERG) for guidance on the initial response. However, when the burning goods are mixed or ...
Highlights- Applied machine learning to identify fire hazards from a combustion analysis dataset.
- Time series classification and augmentations achieve high classification results.
- Self-supervised learning achieves good results without hand-...
Comments
Please enable JavaScript to view thecomments powered by Disqus.Information & Contributors
Information
Published In
August 2023
5996 pages
ISBN:9798400701030
DOI:10.1145/3580305
- General Chairs:
- Ambuj Singh,
- Yizhou Sun,
- Program Chairs:
- Leman Akoglu,
- Dimitrios Gunopulos,
- Xifeng Yan,
- Ravi Kumar,
- Fatma Ozcan,
- Jieping Ye
Copyright © 2023 ACM.
Publication rights licensed to ACM. ACM acknowledges that this contribution was authored or co-authored by an employee, contractor or affiliate of the United States government. As such, the Government retains a nonexclusive, royalty-free right to publish or reproduce this article, or to allow others to do so, for Government purposes only.
Sponsors
Publisher
Association for Computing Machinery
New York, NY, United States
Publication History
Published: 04 August 2023
Check for updates
Author Tags
Qualifiers
- Research-article
Funding Sources
Conference
KDD '23: The 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining
August 6 - 10, 2023
CA, Long Beach, USA
Acceptance Rates
Overall Acceptance Rate 1,133 of 8,635 submissions, 13%
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- 0Total Citations
- 567Total Downloads
- Downloads (Last 12 months)415
- Downloads (Last 6 weeks)27
Reflects downloads up to 18 Nov 2024
Other Metrics
Citations
Cited By
View allView Options
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Sign in