{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/multivariate-time-series-classification-with","title":"Multivariate Time Series Classification with WEASEL+MUSE","arxiv_id":"1711.11343","date":"2017-11-30","proceeding":null,"authors":["Patrick Schäfer","Ulf Leser"],"abstract":"Multivariate time series (MTS) arise when multiple interconnected sensors\nrecord data over time. Dealing with this high-dimensional data is challenging\nfor every classifier for at least two aspects: First, an MTS is not only\ncharacterized by individual feature values, but also by the interplay of\nfeatures in different dimensions. Second, this typically adds large amounts of\nirrelevant data and noise. We present our novel MTS classifier WEASEL+MUSE\nwhich addresses both challenges. WEASEL+MUSE builds a multivariate feature\nvector, first using a sliding-window approach applied to each dimension of the\nMTS, then extracts discrete features per window and dimension. The feature\nvector is subsequently fed through feature selection, removing\nnon-discriminative features, and analysed by a machine learning classifier. The\nnovelty of WEASEL+MUSE lies in its specific way of extracting and filtering\nmultivariate features from MTS by encoding context information into each\nfeature. Still the resulting feature set is small, yet very discriminative and\nuseful for MTS classification. Based on a popular benchmark of 20 MTS datasets,\nwe found that WEASEL+MUSE is among the most accurate classifiers, when compared\nto the state of the art. The outstanding robustness of WEASEL+MUSE is further\nconfirmed based on motion gesture recognition data, where it out-of-the-box\nachieved similar accuracies as domain-specific methods.","url_abs":"http://arxiv.org/abs/1711.11343v4","url_pdf":"http://arxiv.org/pdf/1711.11343v4.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"multivariate-time-series-classification-with","repo_url":"https://github.com/patrickzib/SFA","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"GPL-3.0"}}],"tasks":[{"task_slug":"classification-1","task_name":"Classification"},{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"gesture-recognition","task_name":"Gesture Recognition"},{"task_slug":"time-series-1","task_name":"Time Series"},{"task_slug":"time-series","task_name":"Time Series Analysis"},{"task_slug":"time-series-classification","task_name":"Time Series Classification"},{"task_slug":"feature-selection","task_name":"feature selection"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/time-series-classification-on-aatld-gesture","task":"Time Series Classification","dataset":"AATLD Gesture Recognition","model":"MUSE","rank_in_archive_order":1,"of":1,"metrics":{"Absolute Time (ms)":"133656"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1711.11343","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}