skip to main content
10.1145/3583780.3615082acmconferencesArticle/Chapter ViewAbstractPublication PagescikmConference Proceedingsconference-collections
research-article

Time-series Shapelets with Learnable Lengths

Published: 21 October 2023 Publication History

Abstract

Shapelets are subsequences that are effective for classifying time-series instances. Learning shapelets by a continuous optimization has recently been studied to improve computational efficiency and classification performance. However, existing methods have employed predefined and fixed shapelet lengths during the continuous optimization, despite the fact that shapelets and their lengths are inherently interdependent and thus should be jointly optimized. To efficiently explore shapelets of high quality in terms of interpretability and inter-class separability, this study makes the shapelet lengths continuous and learnable. The proposed formulation jointly optimizes not only a binary classifier and shapelets but also shapelet lengths. The derived SGD optimization can be theoretically interpreted as improving the quality of shapelets in terms of shapelet closeness to the time series for target / off-target classes. We demonstrate improvements in area under the curve, total training time, and shapelet interpretability on UCR binary datasets.

Supplementary Material

MP4 File (full1531.mp4)
Presentation video

References

[1]
Hussein El Amouri, Thomas Andrew Lampert, Pierre Gancc arski, and Clé ment Mallet. 2022. CDPS: Constrained DTW-Preserving Shapelets. In ECML PKDD. Springer, 21--37.
[2]
Anthony J. Bagnall, Jason Lines, Aaron Bostrom, James Large, and Eamonn J. Keogh. 2017. The great time series classification bake off: a review and experimental evaluation of recent algorithmic advances. Data Min. Knowl. Discov., Vol. 31, 3 (2017), 606--660.
[3]
Jo ao Bento, Pedro Saleiro, André F. Cruz, Mário A.T. Figueiredo, and Pedro Bizarro. 2021. TimeSHAP: Explaining Recurrent Models through Sequence Perturbations. In KDD. ACM, 2565--2573.
[4]
Nestor Cabello, Elham Naghizade, Jianzhong Qi, and Lars Kulik. 2020. Fast and Accurate Time Series Classification Through Supervised Interval Search. In ICDM. IEEE Computer Society, 948--953.
[5]
Ziqiang Cheng, Yang Yang, Wei Wang, Wenjie Hu, Yueting Zhuang, and Guojie Song. 2020. Time2Graph: Revisiting Time Series Modeling with Dynamic Shapelets. In AAAI. AAAI Press, 3617--3624.
[6]
Jonathan Crabbé and Mihaela van der Schaar. 2021. Explaining Time Series Predictions with Dynamic Masks. In ICML. PMLR, 2166--2177.
[7]
Marco Cuturi and Mathieu Blondel. 2017. Soft-DTW: A Differentiable Loss Function for Time-Series. In ICML. PMLR, 894--903.
[8]
Hoang Anh Dau, Eamonn Keogh, Kaveh Kamgar, Chin-Chia Yeh, Michael, Yan Zhu, Shaghayegh Gharghabi, Chotirat Ann Ratanamahatana, Yanping, Bing Hu, Nurjahan Begum, Anthony Bagnall, Abdullah Mueen, Gustavo Batista, and Hexagon-ML. 2018. The UCR Time Series Classification Archive. https://www.cs.ucr.edu/ eamonn/time_series_data_2018/.
[9]
Angus Dempster, Franccois Petitjean, and Geoffrey I. Webb. 2020. ROCKET: Exceptionally Fast and Accurate Time Series Classification Using Random Convolutional Kernels. Data Min. Knowl. Discov., Vol. 34, 5 (2020), 1454--1495.
[10]
Angus Dempster, Daniel F. Schmidt, and Geoffrey I. Webb. 2021. MiniRocket: A Very Fast (Almost) Deterministic Transform for Time Series Classification. In KDD. ACM, 248--257.
[11]
Janez Demvsar. 2006. Statistical Comparisons of Classifiers over Multiple Data Sets. J. Mach. Learn. Res., Vol. 7 (2006), 1--30.
[12]
Hui Ding, Goce Trajcevski, Peter Scheuermann, Xiaoyue Wang, and Eamonn Keogh. 2008. Querying and Mining of Time Series Data: Experimental Comparison of Representations and Distance Measures. Proc. VLDB Endow., Vol. 1, 2 (2008), 1542--1552.
[13]
Ramesh Doddaiah, Prathyush S. Parvatharaju, Elke A. Rundensteiner, and Thomas Hartvigsen. 2022. Class-Specific Explainability for Deep Time Series Classifiers. In ICDM. IEEE Computer Society, 101--110.
[14]
Mengnan Du, Ninghao Liu, and Xia Hu. 2019. Techniques for Interpretable Machine Learning. Commun. ACM, Vol. 63, 1 (2019), 68--77.
[15]
Rong-En Fan, Kai-Wei Chang, Cho-Jui Hsieh, Xiang-Rui Wang, and Chih-Jen Lin. 2008. LIBLINEAR: A Library for Large Linear Classification. J. Mach. Learn. Res., Vol. 9 (2008), 1871--1874.
[16]
Zicheng Fang, Peng Wang, and Wei Wang. 2018. Efficient Learning Interpretable Shapelets for Accurate Time Series Classification. In ICDE. IEEE Computer Society, 497--508.
[17]
Len Feremans, Boris Cule, and Bart Goethals. 2022. PETSC: pattern-based embedding for time series classification. Data Min. Knowl. Discov., Vol. 36, 3 (2022), 1015--1061.
[18]
Bryce Goodman and Seth R. Flaxman. 2017. European Union Regulations on Algorithmic Decision-Making and a "Right to Explanation". AI Mag., Vol. 38, 3 (2017), 50--57.
[19]
Josif Grabocka, Nicolas Schilling, Martin Wistuba, and Lars Schmidt-Thieme. 2014. Learning Time-series Shapelets. In KDD. ACM, 392--401.
[20]
Josif Grabocka, Martin Wistuba, and Lars Schmidt-Thieme. 2016. Fast classification of univariate and multivariate time series through shapelet discovery. Knowl. Inf. Syst., Vol. 49, 2 (2016), 429--454.
[21]
Shizhong Han, Zibo Meng, Zhiyuan Li, James O'Reilly, Jie Cai, Xiaofeng Wang, and Yan Tong. 2018. Optimizing Filter Size in Convolutional Neural Networks for Facial Action Unit Recognition. In CVPR. IEEE Computer Society, 5070--5078.
[22]
Jon Hills, Jason Lines, Edgaras Baranauskas, James Mapp, and Anthony J. Bagnall. 2014. Classification of time series by shapelet transformation. Data Min. Knowl. Discov., Vol. 28, 4 (2014), 851--881.
[23]
Lu Hou, James T. Kwok, and Jacek M. Zurada. 2016. Efficient Learning of Timeseries Shapelets. In AAAI. AAAI Press, 1209--1215.
[24]
Tsung-Yu Hsieh, Suhang Wang, Yiwei Sun, and Vasant Honavar. 2021. Explainable Multivariate Time Series Classification: A Deep Neural Network Which Learns to Attend to Important Variables As Well As Time Intervals. In WSDM. ACM, 607--615.
[25]
Hassan Ismail Fawaz, Germain Forestier, Jonathan Weber, Lhassane Idoumghar, and Pierre-Alain Muller. 2019. Deep Learning for Time Series Classification: A Review. Data Min. Knowl. Discov., Vol. 33, 4 (2019), 917--963.
[26]
Hassan Ismail Fawaz, Benjamin Lucas, Germain Forestier, Charlotte Pelletier, Daniel F. Schmidt, Jonathan Weber, Geoffrey I. Webb, Lhassane Idoumghar, Pierre-Alain Muller, and Franccois Petitjean. 2020. InceptionTime: Finding AlexNet for Time Series Classification. Data Min. Knowl. Discov., Vol. 34, 6 (2020), 1936--1962.
[27]
Isak Karlsson, Panagiotis Papapetrou, and Henrik Boströ m. 2016. Generalized random shapelet forests. Data Min. Knowl. Discov., Vol. 30, 5 (2016), 1053--1085.
[28]
Eamonn Keogh and Thanawin Rakthanmanon. 2013. Fast Shapelets: A Scalable Algorithm for Discovering Time Series Shapelets. In SDM. SIAM, 668--676.
[29]
Eamonn J. Keogh and Michael J. Pazzani. 2001. Derivative Dynamic Time Warping. In SDM. SIAM, 1--11.
[30]
Xuan-May Thi Le, Minh-Tuan Tran, and Van-Nam Huynh. 2022. Learning Perceptual Position-Aware Shapelets for Time Series Classification. In ECML PKDD. Springer, 53--69.
[31]
Guozhong Li, Byron Choi, Jianliang Xu, Sourav S. Bhowmick, Kwok-Pan Chun, and Grace Lai-Hung Wong. 2021. ShapeNet: A Shapelet-Neural Network Approach for Multivariate Time Series Classification. In AAAI. AAAI Press, 8375--8383.
[32]
Guozhong Li, Byron Choi, Jianliang Xu, Sourav S. Bhowmick, Daphne Ngar-yin Mah, and Grace Lai-Hung Wong. 2022. IPS: Instance Profile for Shapelet Discovery for Time Series Classification. In ICDE. IEEE Computer Society, 1781--1793.
[33]
Xiaosheng Li and Jessica Lin. 2018. Evolving Separating References for Time Series Classification. In SDM. SIAM, 243--251.
[34]
Jessica Lin, Eamonn J. Keogh, Li Wei, and Stefano Lonardi. 2007. Experiencing SAX: a novel symbolic representation of time series. Data Min. Knowl. Discov., Vol. 15, 2 (2007), 107--144.
[35]
Jason Lines, Luke M. Davis, Jon Hills, and Anthony Bagnall. 2012. A Shapelet Transform for Time Series Classification. In KDD. ACM, 289--297.
[36]
Jason Lines, Sarah Taylor, and Anthony Bagnall. 2018. Time Series Classification with HIVE-COTE: The Hierarchical Vote Collective of Transformation-Based Ensembles. ACM Trans. Knowl. Discov. Data, Vol. 12, 5 (2018), 1--35.
[37]
Jason Lines, Sarah Taylor, and Anthony J. Bagnall. 2016. HIVE-COTE: The Hierarchical Vote Collective of Transformation-Based Ensembles for Time Series Classification. In ICDM. IEEE Computer Society, 1041--1046.
[38]
Scott M. Lundberg and Su-In Lee. 2017. A Unified Approach to Interpreting Model Predictions. In International Conference on Neural Information Processing Systems. Curran Associates Inc., 4768--4777.
[39]
Q. Ma, W. Zhuang, and G. Cottrell. 2019. Triple-Shapelet Networks for Time Series Classification. In ICDM. IEEE Computer Society, 1246--1251.
[40]
Qianli Ma, Wanqing Zhuang, Sen Li, Desen Huang, and G. Cottrell. 2020. Adversarial Dynamic Shapelet Networks. In AAAI. AAAI Press, 5069--5076.
[41]
Pierre-Franccois Marteau. 2009. Time Warp Edit Distance with Stiffness Adjustment for Time Series Matching. IEEE Trans. Pattern Anal. Mach. Intell., Vol. 31, 2 (2009), 306--318.
[42]
Eiji Matsumoto, Kazunori Uchida, Minoru Saito, Akihiro Yamaguchi, and Toshihiro Maekawa. 2022. Recent Digitization of GIS and Sophistication of Equipment Condition Monitoring and Diagnosis applying AI Technologies. In CIGRE Paris. CIGRE, No.10644.
[43]
Matthew Middlehurst, James Large, Michael Flynn, Jason Lines, Aaron Bostrom, and Anthony Bagnall. 2021. HIVE-COTE 2.0: A New Meta Ensemble for Time Series Classification. Mach. Learn., Vol. 110, 11--12 (2021), 3211--3243.
[44]
Abdullah Mueen, Eamonn Keogh, and Neal Young. 2011. Logical-shapelets: An Expressive Primitive for Time Series Classification. In KDD. ACM, 1154--1162.
[45]
Thach Le Nguyen, Severin Gsponer, and Georgiana Ifrim. 2017. Time Series Classification by Sequence Learning in All-Subsequence Space. In ICDE. IEEE Computer Society, 947--958.
[46]
Thach Le Nguyen, Severin Gsponer, Iulia Ilie, Martin O'Reilly, and Georgiana Ifrim. 2019. Interpretable time series classification using linear models and multi-resolution multi-domain symbolic representations. Data Min. Knowl. Discov., Vol. 33, 4 (2019), 1183--1222.
[47]
Prathyush S. Parvatharaju, Ramesh Doddaiah, Thomas Hartvigsen, and Elke A. Rundensteiner. 2021. Learning Saliency Maps to Explain Deep Time Series Classifiers. In CIKM. ACM, 1406--1415.
[48]
Marco Tulio Ribeiro, Sameer Singh, and Carlos Guestrin. 2016. "Why Should I Trust You?": Explaining the Predictions of Any Classifier. In KDD. ACM, 1135--1144.
[49]
Shoumik Roychoudhury, Mohamed Ghalwash, and Zoran Obradovic. 2017. Cost Sensitive Time-Series Classification. In ECML PKDD. Springer, 495--511.
[50]
Shoumik Roychoudhury, Fang Zhou, and Zoran Obradovic. 2019. Leveraging Subsequence-orders for Univariate and Multivariate Time-series Classification. In SDM. SIAM, 495--503.
[51]
Shoumik Roychoudhury, Fang Zhou, and Zoran Obradovic. 2022. Leveraging Dependencies among Learned Temporal Subsequences. In SDM. SIAM, 504--512.
[52]
Cynthia Rudin. 2019. Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nat. Mach. Intell., Vol. 1, 5 (2019), 206--215.
[53]
Patrick Sch"a fer. 2015. The BOSS is concerned with time series classification in the presence of noise. Data Min. Knowl. Discov., Vol. 29, 6 (2015), 1505--1530.
[54]
Patrick Sch"a fer and Mikael Hö gqvist. 2012. SFA: a symbolic fourier approximation and index for similarity search in high dimensional datasets. In EDBT. ACM, 516^^e2^^80^^93527.
[55]
Patrick Sch"a fer and Ulf Leser. 2017. Fast and Accurate Time Series Classification with WEASEL. In CIKM. ACM, 637--646.
[56]
Pavel Senin and Sergey Malinchik. 2013. SAX-VSM: Interpretable Time Series Classification Using SAX and Vector Space Model. In ICDM. IEEE Computer Society, 1175--1180.
[57]
Mit Shah, Josif Grabocka, Nicolas Schilling, Martin Wistuba, and Lars Schmidt-Thieme. 2016. Learning DTW-Shapelets for Time-Series Classification. In CODS. ACM, 1--8.
[58]
Torty Sivill and Peter Flach. 2022. LIMESegment: Meaningful, Realistic Time Series Explanations. In AISTATS. PMLR, 3418--3433.
[59]
Alexandra Stefan, Vassilis Athitsos, and Gautam Das. 2013. The Move-Split-Merge Metric for Time Series. IEEE Trans. Knowl. Data Eng., Vol. 25, 6 (2013), 1425--1438.
[60]
Chang Wei Tan, Angus Dempster, Christoph Bergmeir, and Geoffrey I. Webb. 2022. MultiRocket: multiple pooling operators and transformations for fast and effective time series classification. Data Min. Knowl. Discov., Vol. 36, 5 (2022), 1623--1646.
[61]
Wensi Tang, Lu Liu, and Guodong Long. 2020. Interpretable Time-series Classification on Few-shot Samples. In IJCNN. IEEE Computer Society, 1--8.
[62]
Sana Tonekaboni, Shalmali Joshi, Kieran Campbell, David Duvenaud, and Anna Goldenberg. 2020. What went wrong and when? Instance-wise feature importance for time-series black-box models. In NeurIPS, Vol. 33. Curran Associates, Inc., 799--809.
[63]
Michail Vlachos, Marios Hadjieleftheriou, Dimitrios Gunopulos, and Eamonn J. Keogh. 2006. Indexing Multidimensional Time-Series. VLDB J., Vol. 15, 1 (2006), 1--20.
[64]
Haishuai Wang, Jia Wu, Peng Zhang, and Yixin Chen. 2019a. Learning Shapelet Patterns from Network-Based Time Series. IEEE Trans. Ind. Informatics, Vol. 15, 7 (2019), 3864--3876.
[65]
Haishuai Wang, Qin Zhang, Jia Wu, Shirui Pan, and Yixin Chen. 2019b. Time series feature learning with labeled and unlabeled data. Pattern Recognit., Vol. 89 (2019), 55--66.
[66]
Akihiro Yamaguchi, Shigeru Maya, Kohei Maruchi, and Ken Ueno. 2020b. LTSpAUC: Learning Time-series Shapelets for Optimizing Partial AUC. In SDM. SIAM, 1--9.
[67]
Akihiro Yamaguchi, Shigeru Maya, and Ken Ueno. 2020a. RLTS: Robust Learning Time-series Shapelets. In ECML PKDD. Springer, 595--611.
[68]
Akihiro Yamaguchi and Ken Ueno. 2021. Learning Time-series Shapelets via Supervised Feature Selection. In SDM. SIAM, 262--270.
[69]
Akihiro Yamaguchi, Ken Ueno, and Hisashi Kashima. 2022a. Learning Evolvable Time-series Shapelets. In ICDE. IEEE Computer Society, 793--805.
[70]
Akihiro Yamaguchi, Ken Ueno, and Hisashi Kashima. 2022b. Learning Time-series Shapelets Enhancing Discriminability. In SDM. SIAM, 190--198.
[71]
Akihiro Yamaguchi, Ken Ueno, Kazunori Uchida, Eiji Matsumoto, and Toshiyuki Saida. 2022c. Development of advanced AI technologies for condition diagnosis of high voltage switchgear in substations. CIGRE Science and Engineering, Vol. CSE 026 (2022), 1--11.
[72]
Lexiang Ye and Eamonn Keogh. 2009. Time Series Shapelets: A New Primitive for Data Mining. In KDD. ACM, 947--956.
[73]
Jidong Yuan, Qianhong Lin, Wei Zhang, and Zhihai Wang. 2019. Locally Slope-Based Dynamic Time Warping for Time Series Classification. In CIKM. ACM, 1713--1722.
[74]
Qin Zhang, Jia Wu, Hong Yang, Yingjie Tian, and Chengqi Zhang. 2016. Unsupervised Feature Learning from Time Series. In IJCAI. AAAI Press, 2322--2328.
[75]
Q. Zhang, J. Wu, P. Zhang, G. Long, and C. Zhang. 2018. Salient Subsequence Learning for Time Series Clustering. IEEE Transactions on Pattern Analysis & Machine Intelligence (2018), 2193--2207.
[76]
Han Zou, Yuxun Zhou, Jianfei Yang, Weixi Gu, Lihua Xie, and Costas J. Spanos. 2018. WiFi-Based Human Identification via Convex Tensor Shapelet Learning. In AAAI. AAAI Press, 1711--1719.

Recommendations

Comments

Information & Contributors

Information

Published In

cover image ACM Conferences
CIKM '23: Proceedings of the 32nd ACM International Conference on Information and Knowledge Management
October 2023
5508 pages
ISBN:9798400701245
DOI:10.1145/3583780
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected].

Sponsors

Publisher

Association for Computing Machinery

New York, NY, United States

Publication History

Published: 21 October 2023

Permissions

Request permissions for this article.

Check for updates

Author Tags

  1. classification
  2. interpretability
  3. shapelets
  4. time series

Qualifiers

  • Research-article

Conference

CIKM '23
Sponsor:

Acceptance Rates

Overall Acceptance Rate 1,861 of 8,427 submissions, 22%

Upcoming Conference

CIKM '25

Contributors

Other Metrics

Bibliometrics & Citations

Bibliometrics

Article Metrics

  • 0
    Total Citations
  • 345
    Total Downloads
  • Downloads (Last 12 months)198
  • Downloads (Last 6 weeks)9
Reflects downloads up to 18 Jan 2025

Other Metrics

Citations

View Options

Login options

View options

PDF

View or Download as a PDF file.

PDF

eReader

View online with eReader.

eReader

Media

Figures

Other

Tables

Share

Share

Share this Publication link

Share on social media