[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-83107-en":3,"doc-seo-83107-105":29,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},83107,1099514067438,"River Wang","https://ap-avatar.wpscdn.com/avatar/100002539ee87300030?x-image-process=image/resize,m_fixed,w_180,h_180&k=1780474512215547542",8,"Research & Report","RMISC: A Large-scale Real-world Multivariate Corpus for Time Series Foundation Models","Multivariate time series foundation models (TSFMs) have shown strong zero-shot generalization, yet many are pretrained on multivariate synthetic data that can miss real-world temporal dynamics and cross-variable dependencies. RMISC addresses this gap by introducing an openly accessible, real-world multivariate time series corpus with about 200 datasets and 142B time points across diverse domains. Four TSFMs are pretrained on univariate, synthetic multivariate, and real-world multivariate data, then evaluated on in- and out-of-distribution benchmarks. Results show real-world multivariate pretraining consistently improves generalization for both univariate and multivariate TSFMs.","arXiv :2607 .06504v 1 [ cs .AI ] 7 Jul 2026  \nRMISC: A Large-scale Real-world Multivariate Corpus for Time Series  \nFoundation Models  \nQian Sun 1,2,4,* Yong-Ming Tian 1,2,4,* Jia-Wei Huang 1,2,4 Cheng Feng3,4 Shao-Qun Zhang 1,2,4, B  \n1 State Key Laboratory of Novel Software Technology, Nanjing University, Nanjing, China  \n2 School of Intelligent Science and Technology, Nanjing University, Suzhou, China  \n3 Siemens Data and AIResearch, Beijing, China  \n4 Nanjing University – Siemens Joint Research Center on Industrial AI, Suzhou, China  \nAbstract  \nRecent years have witnessed the emergence of multivariate modeling using time series foundation models (TSFMs), which achieve advanced zero-shot generalization. Modern multivariate TSFMs are predominantly pretrained on multivariate synthetic data, which is easier to scale but may fail to capture the complex temporal dynamics and cross-variable relationships present in real-world time series. This raises a key question: Whether and to what extent the leading TSFMs trained with the real-world corpus perform better than those trained with synthetic data? To answer this, we establish the RMISC corpus, a considerably large-scale, high-quality, openly accessible, real-world, and multivariate time series archive that contains around 200 datasets and 142 billion time points across diverse domains. Furthermore, we pretrain four advanced TSFMs on univariate, synthetic multivariate, and real-world multivariate data and evaluate their zero-shot generalization capabilities on standard in-distribution and out-of-distribution benchmarks. Experimental results show that incorporating real-world multivariate data predominantly improves the generalization performance for both univariate and multivariate TSFMs. These results provide a deeper understanding of how real-world multivariate data contributes to the development of stronger TSFMs.  \nKey words: multivariate time series forecasting, time series foundation model, real-world time series corpus, covariates, out-of-distribution generalization  \n1. Introduction  \nRecent advances in Time Series Foundation Models (TSFMs) have significantly remodeled the paradigm of time series analysis [1] . Fed into large-scale and heterogeneous time series corpora, TSFMs can be directly compatible with diverse forecasting tasks, frequency distributions, and data modalities [2] with remarkable zero-shot generalization capabilities, thus moving beyond traditional statistical methods [3, 4] and deep learning  \nPreprint July 8, 2026  \nMultivariate Univariate  \n\n|  | Univariate Time Series\u003Cbr>􀢞 􀫚: Target \u003Cbr>􀢞 􀫛: Target   \u003Cbr> |  |  | Input Embedding Univariate Models Forecasting Results\u003Cbr> |  |  |  |  |  |\n| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |\n|  | Plenty of Real-World Data Patching  Time Attention Limited Performance |  |  |  |  |  |  |  |  |\n|  |  | Target-Covariate\u003Cbr>Structuring |  |  |  |  | Target-Covariate\u003Cbr>Modeling |  |  |\n|  | Multivariate Time Series\u003Cbr>\u003Cbr>\u003Cbr>\u003Cbr>\u003Cbr>\u003Cbr>􀢞 􀫚 , 􀫚: Target 1\u003Cbr>􀢞 􀫚 ,􀫛 : Target 2\u003Cbr>􀢞 􀫚 ,􀫜 : Covariate\u003Cbr>􀢞 􀫛 , 􀫚: Target 1 \u003Cbr>􀢞 􀫛 ,􀫛 : Target 2  |  |  |  | Input Embedding Multivariate Models Forecasting Results |  |  |  |  |\n|  |  |  |  |  |  |  |  | 􀜡 􀫚 , 􀫚\u003Cbr>􀜡 􀫚 ,􀫛\u003Cbr>􀜡 􀫚 ,􀫜 |  |\n|  |  |  |  |  |  |  |  | \u003Cbr>􀢟 􀫛 , 􀫚 \u003Cbr>\u003Cbr>􀢟 􀫛 ,􀫛 \u003Cbr>􀢟 􀫛 ,􀫜  |  |\n|  | 􀢞 􀫛 ,􀫜 : Covariate |  |  |  |  |  |  |  |  |\n|  |  Time Attention\u003Cbr>Limited to Synthetic Data Patching Strong Performance\u003Cbr>  Group Attention  |  |  |  |  |  |  |  |  |\n\nFigure 1: The modeling workflow of univariate and multivariate time series foundation models on corpora.  \nmodels [5, 6, 7] that repeatedly train task-specific models for individual time series [8, 9] . In recent years, developers have widely applied TSFM to various fields, such as industrial sensing [10], financial assessment [11], healthcare monitoring [12], climate modeling [13], energy management [14], and traffic prediction [15] .  \nCapturing the cross-variable information is o","cbCaiufPvJ6OaMjW","https://ap.wps.com/l/cbCaiufPvJ6OaMjW","pdf",1136636,1,33,"English","en",105,"# Introduction\n## Related Studies","[{\"question\":\"What problem does RMISC aim to solve in time series foundation model training?\",\"answer\":\"It targets the gap between TSFMs pretrained on synthetic multivariate data and the complex temporal dynamics plus cross-variable relationships found in real-world time series.\"},{\"question\":\"What is RMISC, and what scale does it provide?\",\"answer\":\"RMISC is a large-scale real-world multivariate time series corpus that is openly accessible, containing around 200 datasets and about 142 billion time points across diverse domains.\"},{\"question\":\"How do the authors evaluate the impact of using real-world data?\",\"answer\":\"They pretrain four TSFMs on univariate, synthetic multivariate, and real-world multivariate data, then test zero-shot generalization on standard in-distribution and out-of-distribution benchmarks.\"}]",1784185314,83,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":27},"rmisc-a-large-scale-real-world-multivariate-corpus-for-time-series-foundation-models","",{"@graph":35,"@context":85},[36,53,68],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/rmisc-a-large-scale-real-world-multivariate-corpus-for-time-series-foundation-models/83107/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What problem does RMISC aim to solve in time series foundation model training?","Question",{"text":75,"@type":76},"It targets the gap between TSFMs pretrained on synthetic multivariate data and the complex temporal dynamics plus cross-variable relationships found in real-world time series.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What is RMISC, and what scale does it provide?",{"text":80,"@type":76},"RMISC is a large-scale real-world multivariate time series corpus that is openly accessible, containing around 200 datasets and about 142 billion time points across diverse domains.",{"name":82,"@type":73,"acceptedAnswer":83},"How do the authors evaluate the impact of using real-world data?",{"text":84,"@type":76},"They pretrain four TSFMs on univariate, synthetic multivariate, and real-world multivariate data, then test zero-shot generalization on standard in-distribution and out-of-distribution benchmarks.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":45,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":45,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":45,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":45,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":45,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]