{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,1,9]],"date-time":"2026-01-09T20:13:39Z","timestamp":1767989619784,"version":"3.49.0"},"reference-count":44,"publisher":"Springer Science and Business Media LLC","issue":"3","license":[{"start":{"date-parts":[[2024,2,9]],"date-time":"2024-02-09T00:00:00Z","timestamp":1707436800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2024,2,9]],"date-time":"2024-02-09T00:00:00Z","timestamp":1707436800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/100006108","name":"National Center for Advancing Translational Sciences","doi-asserted-by":"publisher","award":["UL1TR003015\/ KL2TR003016"],"award-info":[{"award-number":["UL1TR003015\/ KL2TR003016"]}],"id":[{"id":"10.13039\/100006108","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Data Min Knowl Disc"],"published-print":{"date-parts":[[2024,5]]},"abstract":"<jats:title>Abstract<\/jats:title><jats:p>The multivariate time series classification (MTSC) task aims to predict a class label for a given time series. Recently, modern deep learning-based approaches have achieved promising performance over traditional methods for MTSC tasks. The success of these approaches relies on access to the massive amount of labeled data (i.e., annotating or assigning tags to each sample that shows its corresponding category). However, obtaining a massive amount of labeled data is usually very time-consuming and expensive in many real-world applications such as medicine, because it requires domain experts\u2019 knowledge to annotate data. Insufficient labeled data prevents these models from learning discriminative features, resulting in poor margins that reduce generalization performance. To address this challenge, we propose a novel approach: supervised contrastive learning for time series classification (SupCon-TSC). This approach improves the classification performance by learning the discriminative low-dimensional representations of multivariate time series, and its end-to-end structure allows for interpretable outcomes. It is based on supervised contrastive (SupCon) loss to learn the inherent structure of multivariate time series. First, two separate augmentation families, including strong and weak augmentation methods, are utilized to generate augmented data for the source and target networks, respectively. Second, we propose the instance-level, and cluster-level SupCon learning approaches to capture contextual information to learn the discriminative and universal representation for multivariate time series datasets. In the instance-level SupCon learning approach, for each given anchor instance that comes from the source network, the low-variance output encodings from the target network are sampled as positive and negative instances based on their labels. However, the cluster-level approach is performed between each instance and cluster centers among batches, as opposed to the instance-level approach. The cluster-level SupCon loss attempts to maximize the similarities between each instance and cluster centers among batches. We tested this novel approach on two small cardiopulmonary exercise testing (CPET) datasets and the real-world UEA Multivariate time series archive. The results of the SupCon-TSC model on CPET datasets indicate its capability to learn more discriminative features than existing approaches in situations where the size of the dataset is small. Moreover, the results on the UEA archive show that training a classifier on top of the universal representation features learned by our proposed method outperforms the state-of-the-art approaches.<\/jats:p>","DOI":"10.1007\/s10618-024-01006-1","type":"journal-article","created":{"date-parts":[[2024,2,9]],"date-time":"2024-02-09T09:02:32Z","timestamp":1707469352000},"page":"1493-1519","update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":6,"title":["Universal representation learning for multivariate time series using the instance-level and cluster-level supervised contrastive learning"],"prefix":"10.1007","volume":"38","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-3881-8599","authenticated-orcid":false,"given":"Nazanin","family":"Moradinasab","sequence":"first","affiliation":[]},{"given":"Suchetha","family":"Sharma","sequence":"additional","affiliation":[]},{"given":"Ronen","family":"Bar-Yoseph","sequence":"additional","affiliation":[]},{"given":"Shlomit","family":"Radom-Aizik","sequence":"additional","affiliation":[]},{"given":"Kenneth","family":"C. Bilchick","sequence":"additional","affiliation":[]},{"given":"Dan","family":"M. Cooper","sequence":"additional","affiliation":[]},{"given":"Arthur","family":"Weltman","sequence":"additional","affiliation":[]},{"given":"Donald E.","family":"Brown","sequence":"additional","affiliation":[]}],"member":"297","published-online":{"date-parts":[[2024,2,9]]},"reference":[{"key":"1006_CR1","doi-asserted-by":"crossref","unstructured":"Assaf R, Giurgiu I, Bagehorn F, et\u00a0al (2019) Mtex-cnn: multivariate time series explanations for predictions with convolutional neural networks. In: 2019 IEEE international conference on data mining (ICDM), IEEE, pp 952\u2013957","DOI":"10.1109\/ICDM.2019.00106"},{"key":"1006_CR2","unstructured":"Bagnall A, Dau HA, Lines J, et\u00a0al (2018) The UEA multivariate time series classification archive, 2018. arXiv preprint arXiv:1811.00075"},{"key":"1006_CR3","doi-asserted-by":"publisher","first-page":"596","DOI":"10.1016\/j.ins.2021.05.024","volume":"569","author":"FJ Bald\u00e1n","year":"2021","unstructured":"Bald\u00e1n FJ, Ben\u00edtez JM (2021) Multivariate times series classification through an interpretable representation. Inf Sci 569:596\u2013614","journal-title":"Inf Sci"},{"issue":"15","key":"1006_CR4","volume":"10","author":"R Bar-Yoseph","year":"2022","unstructured":"Bar-Yoseph R, Radom-Aizik S, Coronato N et al (2022) Heart rate and gas exchange dynamic responses to multiple brief exercise bouts (MBEB) in early-and late-pubertal boys and girls. Phys Rep 10(15):e15397","journal-title":"Phys Rep"},{"issue":"2","key":"1006_CR5","doi-asserted-by":"publisher","first-page":"400","DOI":"10.1007\/s10618-014-0349-y","volume":"29","author":"MG Baydogan","year":"2015","unstructured":"Baydogan MG, Runger G (2015) Learning a symbolic representation for multivariate time series classification. Data Min Knowl Discov 29(2):400\u2013422","journal-title":"Data Min Knowl Discov"},{"issue":"2","key":"1006_CR6","doi-asserted-by":"publisher","first-page":"476","DOI":"10.1007\/s10618-015-0425-y","volume":"30","author":"MG Baydogan","year":"2016","unstructured":"Baydogan MG, Runger G (2016) Time series representation and similarity based on local autopatterns. Data Min Knowl Discov 30(2):476\u2013509","journal-title":"Data Min Knowl Discov"},{"issue":"1","key":"1006_CR7","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1186\/s13040-022-00299-6","volume":"15","author":"DE Brown","year":"2022","unstructured":"Brown DE, Sharma S, Jablonski JA et al (2022) Neural network methods for diagnosing patient conditions from cardiopulmonary exercise testing data. BioData Min 15(1):1\u201315","journal-title":"BioData Min"},{"key":"1006_CR8","unstructured":"Chan KP, Fu AWC (1999) Efficient time series matching by wavelets. In: proceedings 15th international conference on data engineering (Cat. No. 99CB36337), IEEE, pp 126\u2013133"},{"key":"1006_CR9","doi-asserted-by":"crossref","unstructured":"Chen L, \u00d6zsu MT, Oria V (2005) Robust and fast similarity search for moving object trajectories. In: proceedings of the 2005 ACM SIGMOD international conference on management of data, pp 491\u2013502","DOI":"10.1145\/1066157.1066213"},{"issue":"12","key":"1006_CR10","doi-asserted-by":"publisher","first-page":"5953","DOI":"10.1109\/JBHI.2022.3206100","volume":"26","author":"N Coronato","year":"2022","unstructured":"Coronato N, Brown DE, Sharma Y et al (2022) Functional data analysis for predicting pediatric failure to complete ten brief exercise bouts. IEEE J Biomed Health Inf 26(12):5953\u20135963","journal-title":"IEEE J Biomed Health Inf"},{"key":"1006_CR11","first-page":"1","volume":"7","author":"J Dem\u0161ar","year":"2006","unstructured":"Dem\u0161ar J (2006) Statistical comparisons of classifiers over multiple data sets. J Mach Learn Res 7:1\u201330","journal-title":"J Mach Learn Res"},{"issue":"1","key":"1006_CR12","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1186\/s12889-020-09154-x","volume":"20","author":"LM Fan","year":"2020","unstructured":"Fan LM, Collins A, Geng L et al (2020) Impact of unhealthy lifestyle on cardiorespiratory fitness and heart rate recovery of medical science students. BMC Public Health 20(1):1\u20138","journal-title":"BMC Public Health"},{"key":"1006_CR13","unstructured":"Fauvel K, Fromont \u00c9, Masson V, et\u00a0al (2020) Local cascade ensemble for multivariate data classification. arXiv preprint arXiv:2005.03645"},{"issue":"23","key":"1006_CR14","doi-asserted-by":"publisher","first-page":"3137","DOI":"10.3390\/math9233137","volume":"9","author":"K Fauvel","year":"2021","unstructured":"Fauvel K, Lin T, Masson V et al (2021) Xcm: an explainable convolutional neural network for multivariate time series classification. Mathematics 9(23):3137","journal-title":"Mathematics"},{"key":"1006_CR15","first-page":"21271","volume":"33","author":"JB Grill","year":"2020","unstructured":"Grill JB, Strub F, Altch\u00e9 F et al (2020) Bootstrap your own latent-a new approach to self-supervised learning. Adv Neural Inf Process Syst 33:21271\u201321284","journal-title":"Adv Neural Inf Process Syst"},{"key":"1006_CR16","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2022.3186963","author":"F He","year":"2022","unstructured":"He F, Ty Fu, Lee WC (2022) Rel-CNN: learning relationship features in time series for classification. IEEE Trans Knowl Data Eng. https:\/\/doi.org\/10.1109\/TKDE.2022.3186963","journal-title":"IEEE Trans Knowl Data Eng"},{"issue":"4","key":"1006_CR17","doi-asserted-by":"publisher","first-page":"917","DOI":"10.1007\/s10618-019-00619-1","volume":"33","author":"H Ismail Fawaz","year":"2019","unstructured":"Ismail Fawaz H, Forestier G, Weber J et al (2019) Deep learning for time series classification: a review. Data Min Knowl Discov 33(4):917\u2013963","journal-title":"Data Min Knowl Discov"},{"issue":"6","key":"1006_CR18","doi-asserted-by":"publisher","first-page":"1936","DOI":"10.1007\/s10618-020-00710-y","volume":"34","author":"H Ismail Fawaz","year":"2020","unstructured":"Ismail Fawaz H, Lucas B, Forestier G et al (2020) Inceptiontime: finding alexnet for time series classification. Data Min Knowl Discov 34(6):1936\u20131962","journal-title":"Data Min Knowl Discov"},{"key":"1006_CR19","doi-asserted-by":"publisher","first-page":"237","DOI":"10.1016\/j.neunet.2019.04.014","volume":"116","author":"F Karim","year":"2019","unstructured":"Karim F, Majumdar S, Darabi H et al (2019) Multivariate LSTM-FCNs for time series classification. Neural Netw 116:237\u2013245","journal-title":"Neural Netw"},{"issue":"5","key":"1006_CR20","doi-asserted-by":"publisher","first-page":"1053","DOI":"10.1007\/s10618-016-0473-y","volume":"30","author":"I Karlsson","year":"2016","unstructured":"Karlsson I, Papapetrou P, Bostr\u00f6m H (2016) Generalized random shapelet forests. Data Min Knowl Discov 30(5):1053\u20131085","journal-title":"Data Min Knowl Discov"},{"issue":"4","key":"1006_CR21","doi-asserted-by":"publisher","first-page":"349","DOI":"10.1023\/A:1024988512476","volume":"7","author":"E Keogh","year":"2003","unstructured":"Keogh E, Kasetty S (2003) On the need for time series data mining benchmarks: a survey and empirical demonstration. Data Min Knowl Discov 7(4):349\u2013371","journal-title":"Data Min Knowl Discov"},{"key":"1006_CR22","first-page":"18661","volume":"33","author":"P Khosla","year":"2020","unstructured":"Khosla P, Teterwak P, Wang C et al (2020) Supervised contrastive learning. Adv Neural Inf Process Syst 33:18661\u201318673","journal-title":"Adv Neural Inf Process Syst"},{"key":"1006_CR23","doi-asserted-by":"crossref","unstructured":"Lines J, Taylor S, Bagnall A (2016) Hive-cote: The hierarchical vote collective of transformation-based ensembles for time series classification. In: 2016 IEEE 16th international conference on data mining (ICDM), IEEE, pp 1041\u20131046","DOI":"10.1109\/ICDM.2016.0133"},{"issue":"3","key":"1006_CR24","doi-asserted-by":"publisher","first-page":"607","DOI":"10.1007\/s10618-019-00617-3","volume":"33","author":"B Lucas","year":"2019","unstructured":"Lucas B, Shifaz A, Pelletier C et al (2019) Proximity forest: an effective and scalable distance-based classifier for time series. Data Min Knowl Discov 33(3):607\u2013635","journal-title":"Data Min Knowl Discov"},{"issue":"11","key":"1006_CR25","doi-asserted-by":"publisher","first-page":"2445","DOI":"10.1007\/s00421-020-04457-8","volume":"120","author":"T Matsuo","year":"2020","unstructured":"Matsuo T, So R, Takahashi M (2020) Estimating cardiorespiratory fitness from heart rates both during and after stepping exercise: a validated simple and safe procedure for step tests at worksites. Eur J Appl Physiol 120(11):2445\u20132454","journal-title":"Eur J Appl Physiol"},{"issue":"11","key":"1006_CR26","doi-asserted-by":"publisher","first-page":"3211","DOI":"10.1007\/s10994-021-06057-9","volume":"110","author":"M Middlehurst","year":"2021","unstructured":"Middlehurst M, Large J, Flynn M et al (2021) Hive-cote 2.0: a new meta ensemble for time series classification. Mach Learn 110(11):3211\u20133243","journal-title":"Mach Learn"},{"key":"1006_CR27","doi-asserted-by":"crossref","unstructured":"Minnen D, Starner T, Essa I, et\u00a0al (2006) Discovering characteristic actions from on-body sensor data. In: 2006 10th IEEE international symposium on wearable computers, IEEE, pp 11\u201318","DOI":"10.1109\/ISWC.2006.286337"},{"key":"1006_CR28","doi-asserted-by":"crossref","unstructured":"Rakthanmanon T, Keogh E (2013) Fast shapelets: a scalable algorithm for discovering time series shapelets. In: proceedings of the 2013 SIAM international conference on data mining, SIAM, pp 668\u2013676","DOI":"10.1137\/1.9781611972832.74"},{"issue":"2","key":"1006_CR29","doi-asserted-by":"publisher","first-page":"401","DOI":"10.1007\/s10618-020-00727-3","volume":"35","author":"AP Ruiz","year":"2021","unstructured":"Ruiz AP, Flynn M, Large J et al (2021) The great multivariate time series classification bake off: a review and experimental evaluation of recent algorithmic advances. Data Min Knowl Discov 35(2):401\u2013449","journal-title":"Data Min Knowl Discov"},{"key":"1006_CR30","doi-asserted-by":"crossref","unstructured":"Sch\u00e4fer P, Leser U (2017) Multivariate time series classification with weasel+ muse. arXiv preprint arXiv:1711.11343","DOI":"10.1145\/3132847.3132980"},{"key":"1006_CR31","doi-asserted-by":"crossref","unstructured":"Selvaraju RR, Cogswell M, Das A, et\u00a0al (2017) Grad-cam: visual explanations from deep networks via gradient-based localization. In: Proceedings of the IEEE international conference on computer vision, pp 618\u2013626","DOI":"10.1109\/ICCV.2017.74"},{"issue":"1\u201323","key":"1006_CR32","first-page":"40","volume":"855","author":"P Senin","year":"2008","unstructured":"Senin P (2008) Dynamic time warping algorithm review. Inf Comput Sci Dep Univ Hawaii at Manoa Honolulu USA 855(1\u201323):40","journal-title":"Inf Comput Sci Dep Univ Hawaii at Manoa Honolulu USA"},{"key":"1006_CR33","doi-asserted-by":"crossref","unstructured":"Senin P, Malinchik S (2013) Sax-vsm: interpretable time series classification using sax and vector space model. In: 2013 IEEE 13th international conference on data mining, IEEE, pp 1175\u20131180","DOI":"10.1109\/ICDM.2013.52"},{"issue":"1","key":"1006_CR34","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1007\/s10618-016-0455-0","volume":"31","author":"M Shokoohi-Yekta","year":"2017","unstructured":"Shokoohi-Yekta M, Hu B, Jin H et al (2017) Generalizing DTW to the multi-dimensional case requires an adaptive approach. Data Min Knowl Discov 31(1):1\u201331","journal-title":"Data Min Knowl Discov"},{"key":"1006_CR35","unstructured":"Sukhbaatar S, Bruna J, Paluri M, et\u00a0al (2014) Training convolutional networks with noisy labels. arXiv preprint arXiv:1406.2080"},{"key":"1006_CR36","doi-asserted-by":"publisher","first-page":"202","DOI":"10.1016\/j.patcog.2017.08.016","volume":"73","author":"KS Tuncel","year":"2018","unstructured":"Tuncel KS, Baydogan MG (2018) Autoregressive forests for multivariate time series modeling. Pattern Recogn 73:202\u2013215","journal-title":"Pattern Recogn"},{"key":"1006_CR37","doi-asserted-by":"crossref","unstructured":"Wang X, Gao Y, Lin J, et\u00a0al (2015) A machine learning approach to false alarm detection for critical arrhythmia alarms. In: 2015 IEEE 14th international conference on machine learning and applications (ICMLA), IEEE, pp 202\u2013207","DOI":"10.1109\/ICMLA.2015.176"},{"key":"1006_CR38","doi-asserted-by":"crossref","unstructured":"Wang Z, Yan W, Oates T (2017) Time series classification from scratch with deep neural networks: A strong baseline. In: 2017 international joint conference on neural networks (IJCNN), IEEE, pp 1578\u20131585","DOI":"10.1109\/IJCNN.2017.7966039"},{"key":"1006_CR39","doi-asserted-by":"crossref","unstructured":"Wang X, Fan H, Tian Y, et\u00a0al (2022) On the importance of asymmetry for siamese representation learning. In: Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition, pp 16570\u201316579","DOI":"10.1109\/CVPR52688.2022.01607"},{"key":"1006_CR40","unstructured":"Wistuba M, Grabocka J, Schmidt-Thieme L (2015) Ultra-fast shapelets for time series classification. arXiv preprint arXiv:1503.05018"},{"key":"1006_CR41","doi-asserted-by":"crossref","unstructured":"Ye L, Keogh E (2009) Time series shapelets: a new primitive for data mining. In: Proceedings of the 15th ACM SIGKDD international conference on knowledge discovery and data mining, pp 947\u2013956","DOI":"10.1145\/1557019.1557122"},{"key":"1006_CR42","doi-asserted-by":"crossref","unstructured":"Yue Z, Wang Y, Duan J, et al (2022) Ts2vec: towards universal representation classification with attentional prototypical network. In: proceedings of the AAAI conference on artificial intelligence, pp 6845\u20136852","DOI":"10.1609\/aaai.v36i8.20881"},{"key":"1006_CR43","unstructured":"Zhang Z, Sabuncu M (2018) Generalized cross entropy loss for training deep neural networks with noisy labels. Adv Neural Inf Process Syst 31"},{"key":"1006_CR44","doi-asserted-by":"crossref","unstructured":"Zhang X, Gao Y, Lin J, et\u00a0al (2020) Tapnet: multivariate time series classification with attentional prototypical network. In: proceedings of the AAAI conference on artificial intelligence, pp 6845\u20136852","DOI":"10.1609\/aaai.v34i04.6165"}],"container-title":["Data Mining and Knowledge Discovery"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10618-024-01006-1.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s10618-024-01006-1\/fulltext.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10618-024-01006-1.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,5,4]],"date-time":"2024-05-04T09:16:41Z","timestamp":1714814201000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s10618-024-01006-1"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,2,9]]},"references-count":44,"journal-issue":{"issue":"3","published-print":{"date-parts":[[2024,5]]}},"alternative-id":["1006"],"URL":"https:\/\/doi.org\/10.1007\/s10618-024-01006-1","relation":{},"ISSN":["1384-5810","1573-756X"],"issn-type":[{"value":"1384-5810","type":"print"},{"value":"1573-756X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,2,9]]},"assertion":[{"value":"26 February 2023","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"2 January 2024","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"9 February 2024","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"Not applicable.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Conflict of interest"}},{"value":"Not applicable.","order":3,"name":"Ethics","group":{"name":"EthicsHeading","label":"Ethical approval"}},{"value":"Not applicable.","order":4,"name":"Ethics","group":{"name":"EthicsHeading","label":"Consent to participate"}},{"value":"Not applicable.","order":5,"name":"Ethics","group":{"name":"EthicsHeading","label":"Consent for publication"}}]}}