{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,19]],"date-time":"2026-06-19T23:15:55Z","timestamp":1781910955678,"version":"3.54.5"},"reference-count":52,"publisher":"Springer Science and Business Media LLC","issue":"5","license":[{"start":{"date-parts":[[2025,8,7]],"date-time":"2025-08-07T00:00:00Z","timestamp":1754524800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2025,8,7]],"date-time":"2025-08-07T00:00:00Z","timestamp":1754524800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/501100001775","name":"University of Technology Sydney","doi-asserted-by":"crossref","id":[{"id":"10.13039\/501100001775","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Data Min Knowl Disc"],"published-print":{"date-parts":[[2025,9]]},"abstract":"<jats:title>Abstract<\/jats:title>\n          <jats:p>Large Language Models (LLMs) are widely applied in time series analysis. Yet, their utility in few-shot classification\u2014a scenario with limited training data\u2014remains unexplored. We aim to leverage the pre-trained knowledge in LLMs to overcome the data scarcity problem within multivariate time series. To this end, we propose <jats:bold>LLMFew<\/jats:bold>, an LLM-enhanced framework, to investigate the feasibility and capacity of LLMs for few-shot multivariate time series classification (MTSC). We first introduce a <jats:bold>P<\/jats:bold>atch-wise <jats:bold>T<\/jats:bold>emporal <jats:bold>C<\/jats:bold>onvolution <jats:bold>Enc<\/jats:bold>oder (PTCEnc) to align time series data with the textual embedding input of LLMs. Then, we fine-tune the pre-trained LLM decoder with Low-rank Adaptations (LoRA) to enable effective representation learning from time series data. Experimental results show our model consistently outperforms state-of-the-art baselines by a large margin, achieving 125.2% and 50.2% improvement in classification accuracy on Handwriting and EthanolConcentration datasets, respectively. Our results also show LLM-based methods achieve comparable performance to traditional models across various datasets in few-shot MTSC, paving the way for applying LLMs in practical scenarios where labeled data are limited. Our code is available at <jats:ext-link xmlns:xlink=\"http:\/\/www.w3.org\/1999\/xlink\" xlink:href=\"https:\/\/github.com\/junekchen\/llm-fewshot-mtsc\" ext-link-type=\"uri\">https:\/\/github.com\/junekchen\/llm-fewshot-mtsc<\/jats:ext-link>.<\/jats:p>","DOI":"10.1007\/s10618-025-01145-z","type":"journal-article","created":{"date-parts":[[2025,8,7]],"date-time":"2025-08-07T04:43:39Z","timestamp":1754541819000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":6,"title":["Large language models are few-shot multivariate time series classifiers"],"prefix":"10.1007","volume":"39","author":[{"given":"Yakun","family":"Chen","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zihao","family":"Li","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chao","family":"Yang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xianzhi","family":"Wang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Guandong","family":"Xu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2025,8,7]]},"reference":[{"key":"1145_CR1","unstructured":"Abdin M, Jacobs SA, Awan AA, Aneja J, Awadallah A, Awadalla H, Bach N, Bahree A, Bakhtiari A, Behl H, et al (2024) Phi-3 technical report: A highly capable language model locally on your phone. Preprint at arXiv:2404.14219"},{"key":"1145_CR2","unstructured":"Bagnall A, Dau HA, Lines J, Flynn M, Large J, Bostrom A, Southam P, Keogh E (2018) The uea multivariate time series classification archive, 2018. Preprint at arXiv:1811.00075"},{"key":"1145_CR3","unstructured":"Bai J, Bai S, Chu Y, Cui Z, Dang K, Deng X, Fan Y, Ge W, Han Y, Huang F et al (2023) Qwen technical report. Preprint at arXiv:2309.16609"},{"key":"1145_CR4","unstructured":"Bai S, Kolter JZ, Koltun V (2018) An empirical evaluation of generic convolutional and recurrent networks for sequence modeling. Preprint atarXiv:1803.01271"},{"key":"1145_CR5","doi-asserted-by":"crossref","unstructured":"Barreda J, Gomez A, Puga R, Zhou K, Zhang L (2024) Cosco: a sharpness-aware training framework for few-shot multivariate time series classification. In: Proceedings of the 33rd ACM international conference on information and knowledge management, pp 3622\u20133626","DOI":"10.1145\/3627673.3679891"},{"key":"1145_CR6","unstructured":"Biderman S, Schoelkopf H, Anthony QG, Bradley H, O\u2019Brien K, Hallahan E, Khan MA, Purohit S, Prashanth US, Raff E, et al (2023) Pythia: a suite for analyzing large language models across training and scaling. In: International conference on machine learning, PMLR, pp 2397\u20132430"},{"key":"1145_CR7","unstructured":"Cao D, Jia F, Arik SO, Pfister T, Zheng Y, Ye W, Liu Y (2023) Tempo: Prompt-based generative pre-trained transformer for time series forecasting. Preprint at arXiv:2310.04948"},{"key":"1145_CR8","doi-asserted-by":"publisher","unstructured":"Cheng M, Chen Y, Liu Q, Liu Z, Luo Y, Chen E (2025) Instructime: advancing time series classification with multimodal language modeling. In: Proceedings of the eighteenth ACM international conference on web search and data mining. WSDM \u201925, pp 792\u2013800. Association for Computing Machinery, New York. https:\/\/doi.org\/10.1145\/3701551.3703499","DOI":"10.1145\/3701551.3703499"},{"key":"1145_CR9","doi-asserted-by":"crossref","unstructured":"Chen X, Ge C, Wang M, Wang J (2023) Supervised contrastive few-shot learning for high-frequency time series. In: Proceedings of the AAAI conference on artificial intelligence, vol 37, pp 7069\u20137077","DOI":"10.1609\/aaai.v37i6.25863"},{"key":"1145_CR10","unstructured":"Das A, Kong W, Sen R, Zhou Y (2024) A decoder-only foundation model for time-series forecasting. In: Forty-first international conference on machine learning"},{"issue":"5","key":"1145_CR11","doi-asserted-by":"publisher","first-page":"1454","DOI":"10.1007\/s10618-020-00701-z","volume":"34","author":"A Dempster","year":"2020","unstructured":"Dempster A, Petitjean F, Webb GI (2020) Rocket: exceptionally fast and accurate time series classification using random convolutional kernels. Data Min Knowl Disc 34(5):1454\u20131495","journal-title":"Data Min Knowl Disc"},{"key":"1145_CR12","unstructured":"Dubey A, Jauhri A, Pandey A, Kadian A, Al-Dahle A, Letman A, Mathur A, Schelten A, Yang A, Fan A, et al (2024) The llama 3 herd of models. Preprint at arXiv:2407.21783"},{"key":"1145_CR13","first-page":"74147","volume":"37","author":"V Ekambaram","year":"2024","unstructured":"Ekambaram V, Jati A, Dayama P, Mukherjee S, Nguyen N, Gifford WM, Reddy C, Kalagnanam J (2024) Tiny time mixers (ttms): fast pre-trained models for enhanced zero\/few-shot forecasting of multivariate time series. Adv Neural Inf Process Syst 37:74147\u201374181","journal-title":"Adv Neural Inf Process Syst"},{"key":"1145_CR14","doi-asserted-by":"crossref","unstructured":"Fawaz HI, Forestier G, Weber J, Idoumghar L, Muller PA (2018) Transfer learning for time series classification. In: 2018 IEEE international conference on big data (Big Data), IEEE, pp 1367\u20131376","DOI":"10.1109\/BigData.2018.8621990"},{"key":"1145_CR15","unstructured":"Finn C, Abbeel P, Levine S (2017) Model-agnostic meta-learning for fast adaptation of deep networks. In: International conference on machine learning, PMLR, pp 1126\u20131135"},{"key":"1145_CR16","unstructured":"Garza A, Challu C, Mergenthaler-Canseco M (2023) Timegpt-1. Preprint at arXiv:2310.03589"},{"key":"1145_CR17","unstructured":"Gruver N, Finzi M, Qiu S, Wilson AG (2024) Large language models are zero-shot time series forecasters. Adv Neural Inf Process Syst 36"},{"key":"1145_CR18","doi-asserted-by":"crossref","unstructured":"Gudmundsson S, Runarsson TP, Sigurdsson S(2008) Support vector machines and dynamic time warping for time series. In: 2008 IEEE international joint conference on neural networks (IEEE World Congress on Computational Intelligence), IEEE, pp 2772\u20132776","DOI":"10.1109\/IJCNN.2008.4634188"},{"key":"1145_CR19","doi-asserted-by":"crossref","unstructured":"Gupta P, Bhaskarpandit S, Gupta M (2021) Similarity learning based few shot learning for ECG time series classification. In: 2021 digital image computing: techniques and applications (DICTA), IEEE, pp 1\u20138","DOI":"10.1109\/DICTA52665.2021.9647357"},{"key":"1145_CR20","unstructured":"Hu EJ, Shen Y, Wallis P, Allen-Zhu Z, Li Y, Wang S, Wang L, Chen W (2021) Lora: low-rank adaptation of large language models. Preprint at arXiv:2106.09685"},{"issue":"4","key":"1145_CR21","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, Idoumghar L, Muller P-A (2019) Deep learning for time series classification: a review. Data Min Knowl Disc 33(4):917\u2013963","journal-title":"Data Min Knowl Disc"},{"issue":"5","key":"1145_CR22","doi-asserted-by":"publisher","first-page":"2653","DOI":"10.1007\/s11280-019-00775-w","volume":"23","author":"J Jiang","year":"2020","unstructured":"Jiang J, Ji S, Long G (2020) Decentralized knowledge acquisition for mobile internet applications. World Wide Web 23(5):2653\u20132669","journal-title":"World Wide Web"},{"key":"1145_CR23","unstructured":"Jin M, Wang S, Ma L, Chu Z, Zhang JY, Shi X, Chen PY, Liang Y, Li YF, Pan S, et al (2023) Time-llm: time series forecasting by reprogramming large language models. Preprint at arXiv:2310.01728"},{"key":"1145_CR24","unstructured":"Jin M, Zhang Y, Chen W, Zhang K, Liang Y, Yang B, Wang J, Pan S, Wen Q (2024) Position: what can large language models tell us about time series analysis. In: Forty-first international conference on machine learning"},{"key":"1145_CR25","doi-asserted-by":"publisher","first-page":"1662","DOI":"10.1109\/ACCESS.2017.2779939","volume":"6","author":"F Karim","year":"2017","unstructured":"Karim F, Majumdar S, Darabi H, Chen S (2017) Lstm fully convolutional networks for time series classification. IEEE Access 6:1662\u20131669","journal-title":"IEEE Access"},{"key":"1145_CR26","doi-asserted-by":"crossref","unstructured":"Kaur R, Zeng Z, Balch T, Veloso M (2024) Lets-c: leveraging language embedding for time series classification. Preprint at arXiv:2407.06533","DOI":"10.18653\/v1\/2025.acl-long.1557"},{"key":"1145_CR27","unstructured":"Kitaev N, Kaiser \u0141, Levskaya A (2020) Reformer: the efficient transformer. Preprint at arXiv:2001.04451"},{"key":"1145_CR28","doi-asserted-by":"crossref","unstructured":"Le XM, Luo L, Aickelin U, Tran MT (2024) Shapeformer: Shapelet transformer for multivariate time series classification. Preprint at arXiv:2405.14608","DOI":"10.1145\/3637528.3671862"},{"key":"1145_CR29","doi-asserted-by":"crossref","unstructured":"Lu P, Gong R, Jiang S, Qiu L, Huang S, Liang X, Zhu SC (2021) Inter-gps: interpretable geometry problem solving with formal language and symbolic reasoning. Preprint at arXiv:2105.04165","DOI":"10.18653\/v1\/2021.acl-long.528"},{"key":"1145_CR30","doi-asserted-by":"crossref","unstructured":"Narwariya J, Malhotra P, Vig L, Shroff G, Vishnu T (2020) Meta-learning for few-shot time series classification. In: Proceedings of the 7th ACM IKDD CoDS and 25th COMAD, pp 28\u201336","DOI":"10.1145\/3371158.3371162"},{"key":"1145_CR31","unstructured":"Nie Y, Nguyen NH, Sinthong P, Kalagnanam J (2022) A time series is worth 64 words: long-term forecasting with transformers. Preprint arXiv:2211.14730"},{"key":"1145_CR32","unstructured":"Pan Z, Jiang Y, Garg S, Schneider A, Nevmyvaka Y, Song D (2024) [CDATA[\\textbf{S}^2]]$$\\textbf{S}^2$$ip-llm: semantic space informed prompt learning with llm for time series forecasting. In: Forty-first international conference on machine learning"},{"key":"1145_CR33","doi-asserted-by":"publisher","first-page":"41400","DOI":"10.1109\/ACCESS.2023.3270493","volume":"11","author":"S-H Park","year":"2023","unstructured":"Park S-H, Syazwany NS, Lee S-C (2023) Meta-feature fusion for few-shot time series classification. IEEE Access 11:41400\u201341414","journal-title":"IEEE Access"},{"issue":"8","key":"1145_CR34","first-page":"9","volume":"1","author":"A Radford","year":"2019","unstructured":"Radford A, Wu J, Child R, Luan D, Amodei D, Sutskever I et al (2019) Language models are unsupervised multitask learners. OpenAI blog 1(8):9","journal-title":"OpenAI blog"},{"issue":"1","key":"1145_CR35","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/s41746-018-0029-1","volume":"1","author":"A Rajkomar","year":"2018","unstructured":"Rajkomar A, Oren E, Chen K, Dai AM, Hajaj N, Hardt M, Liu PJ, Liu X, Marcus J, Sun M et al (2018) Scalable and accurate deep learning with electronic health records. NPJ Digit Med 1(1):1\u201310","journal-title":"NPJ Digit Med"},{"key":"1145_CR36","unstructured":"Snell J, Swersky K, Zemel R (2017) Prototypical networks for few-shot learning. Adv Neural Inf Process Syst 30"},{"key":"1145_CR37","doi-asserted-by":"crossref","unstructured":"Taga EO, Ildiz ME, Oymak S (2025) Timepfn: effective multivariate time series forecasting with synthetic data. In: Proceedings of the AAAI conference on artificial intelligence, vol 39, pp 20761\u201320769","DOI":"10.1609\/aaai.v39i19.34288"},{"key":"1145_CR38","doi-asserted-by":"crossref","unstructured":"Tang W, Liu L, Long G (2020) Interpretable time-series classification on few-shot samples. In: 2020 international joint conference on neural networks (IJCNN), IEEE, pp 1\u20138","DOI":"10.1109\/IJCNN48605.2020.9206860"},{"key":"1145_CR39","unstructured":"Tang W, Long G, Liu L, Zhou T, Jiang J, Blumenstein M (2020) Rethinking 1d-cnn for time series classification: A stronger baseline. Preprint at arXiv:2002.10061, pp 1\u20137"},{"key":"1145_CR40","unstructured":"Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Gomez AN, Kaiser \u0141, Polosukhin I (2017) Attention is all you need. Adv Neural Inf Process Syst 30"},{"issue":"3","key":"1145_CR41","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/3386252","volume":"53","author":"Y Wang","year":"2020","unstructured":"Wang Y, Yao Q, Kwok JT, Ni LM (2020) Generalizing from a few examples: a survey on few-shot learning. ACM Comput Surv (csur) 53(3):1\u201334","journal-title":"ACM Comput Surv (csur)"},{"key":"1145_CR42","first-page":"22419","volume":"34","author":"H Wu","year":"2021","unstructured":"Wu H, Xu J, Wang J, Long M (2021) Autoformer: decomposition transformers with auto-correlation for long-term series forecasting. Adv Neural Inf Process Syst 34:22419\u201322430","journal-title":"Adv Neural Inf Process Syst"},{"key":"1145_CR43","unstructured":"Wu H, Hu T, Liu Y, Zhou H, Wang J, Long M (2022) Timesnet: temporal 2d-variation modeling for general time series analysis. Preprint at arXiv:2210.02186"},{"key":"1145_CR44","doi-asserted-by":"publisher","first-page":"6851","DOI":"10.1109\/TKDE.2023.3342137","volume":"36","author":"H Xue","year":"2023","unstructured":"Xue H, Salim FD (2023) Promptcast: a new prompt-based learning paradigm for time series forecasting. IEEE Trans Knowl Data Eng 36:6851","journal-title":"IEEE Trans Knowl Data Eng"},{"key":"1145_CR45","doi-asserted-by":"publisher","DOI":"10.1016\/j.ins.2023.119881","volume":"656","author":"C Yang","year":"2024","unstructured":"Yang C, Wang X, Yao L, Long G, Xu G (2024) Dyformer: a dynamic transformer-based architecture for multivariate time series classification. Inf Sci 656:119881","journal-title":"Inf Sci"},{"key":"1145_CR46","doi-asserted-by":"crossref","unstructured":"Zeng A, Chen M, Zhang L, Xu Q (2023) Are transformers effective for time series forecasting? In: Proceedings of the AAAI conference on artificial intelligence, vol 37, pp 11121\u201311128","DOI":"10.1609\/aaai.v37i9.26317"},{"key":"1145_CR47","first-page":"3988","volume":"35","author":"X Zhang","year":"2022","unstructured":"Zhang X, Zhao Z, Tsiligkaridis T, Zitnik M (2022) Self-supervised contrastive pre-training for time series via time-frequency consistency. Adv Neural Inf Process Syst 35:3988\u20134003","journal-title":"Adv Neural Inf Process Syst"},{"key":"1145_CR48","doi-asserted-by":"crossref","unstructured":"Zhang H, Pang Z, Wang J, Li T (2023) Few-shot learning using data augmentation and time-frequency transformation for time series classification. In: 2023 5th international conference on robotics, intelligent control and artificial intelligence (RICAI), IEEE, pp 733\u2013738","DOI":"10.1109\/RICAI60863.2023.10489648"},{"key":"1145_CR49","unstructured":"Zhang Y, Yan J (2023) Crossformer: Transformer utilizing cross-dimension dependency for multivariate time series forecasting. In: The eleventh international conference on learning representations"},{"key":"1145_CR50","first-page":"43322","volume":"36","author":"T Zhou","year":"2023","unstructured":"Zhou T, Niu P, Sun L, Jin R et al (2023) One fits all: power general time series analysis by pretrained lm. Adv Neural Inf Process Syst 36:43322\u201343355","journal-title":"Adv Neural Inf Process Syst"},{"key":"1145_CR51","unstructured":"Zhou T, Ma Z, Wen Q, Wang X, Sun L, Jin R (2022) Fedformer: frequency enhanced decomposed transformer for long-term series forecasting. In: International conference on machine learning, PMLR, pp 27268\u201327286"},{"key":"1145_CR52","doi-asserted-by":"crossref","unstructured":"Zhou H, Zhang S, Peng J, Zhang S, Li J, Xiong H, Zhang W(2021) Informer: beyond efficient transformer for long sequence time-series forecasting. In: Proceedings of the AAAI conference on artificial intelligence, vol 35, pp 11106\u201311115","DOI":"10.1609\/aaai.v35i12.17325"}],"container-title":["Data Mining and Knowledge Discovery"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10618-025-01145-z.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s10618-025-01145-z\/fulltext.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10618-025-01145-z.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,9,12]],"date-time":"2025-09-12T10:29:30Z","timestamp":1757672970000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s10618-025-01145-z"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,8,7]]},"references-count":52,"journal-issue":{"issue":"5","published-print":{"date-parts":[[2025,9]]}},"alternative-id":["1145"],"URL":"https:\/\/doi.org\/10.1007\/s10618-025-01145-z","relation":{},"ISSN":["1384-5810","1573-756X"],"issn-type":[{"value":"1384-5810","type":"print"},{"value":"1573-756X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,8,7]]},"assertion":[{"value":"8 June 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"19 July 2025","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"7 August 2025","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"The authors declare no Conflict of interest.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Conflict of interest"}}],"article-number":"66"}}