HIVE-COTE 2.0: a new meta ensemble for time series classification. Middlehurst, M., Large, J., Flynn, M., Lines, J., Bostrom, A., & Bagnall, A. Machine Learning, 110:3211?3243, December, 2021. Funding Information: This work is supported by the UK Engineering and Physical Sciences Research Council (EPSRC) through an iCASE award sponsored by British Telecom (T206188) and an equipment Grant (T024593). The experiments were carried out on the High Performance Computing Cluster supported by the Research and Specialist Computing Support service at the University of East Anglia and using a Titan X Pascal donated by the NVIDIA Corporation.![HIVE-COTE 2.0: a new meta ensemble for time series classification [link]](https://bibbase.org/img/filetypes/link.svg "link")
Paper doi abstract bibtex The Hierarchical Vote Collective of Transformation-based Ensembles (HIVE-COTE) is a heterogeneous meta ensemble for time series classification. HIVE-COTE forms its ensemble from classifiers of multiple domains, including phase-independent shapelets, bag-of-words based dictionaries and phase-dependent intervals. Since it was first proposed in 2016, the algorithm has remained state of the art for accuracy on the UCR time series classification archive. Over time it has been incrementally updated, culminating in its current state, HIVE-COTE 1.0. During this time a number of algorithms have been proposed which match the accuracy of HIVE-COTE. We propose comprehensive changes to the HIVE-COTE algorithm which significantly improve its accuracy and usability, presenting this upgrade as HIVE-COTE 2.0. We introduce two novel classifiers, the Temporal Dictionary Ensemble and Diverse Representation Canonical Interval Forest, which replace existing ensemble members. Additionally, we introduce the Arsenal, an ensemble of ROCKET classifiers as a new HIVE-COTE 2.0 constituent. We demonstrate that HIVE-COTE 2.0 is significantly more accurate on average than the current state of the art on 112 univariate UCR archive datasets and 26 multivariate UEA archive datasets.
@article{uea80910,
note = {Funding Information: This work is supported by the UK Engineering and Physical Sciences Research Council (EPSRC) through an iCASE award sponsored by British Telecom (T206188) and an equipment Grant (T024593). The experiments were carried out on the High Performance Computing Cluster supported by the Research and Specialist Computing Support service at the University of East Anglia and using a Titan X Pascal donated by the NVIDIA Corporation.},
month = {December},
author = {Matthew Middlehurst and James Large and Michael Flynn and Jason Lines and Aaron Bostrom and Anthony Bagnall},
title = {HIVE-COTE 2.0: a new meta ensemble for time series classification},
journal = {Machine Learning},
year = {2021},
doi = {10.1007/s10994-021-06057-9},
volume = {110},
pages = {3211?3243},
abstract = {The Hierarchical Vote Collective of Transformation-based Ensembles (HIVE-COTE) is a heterogeneous meta ensemble for time series classification. HIVE-COTE forms its ensemble from classifiers of multiple domains, including phase-independent shapelets, bag-of-words based dictionaries and phase-dependent intervals. Since it was first proposed in 2016, the algorithm has remained state of the art for accuracy on the UCR time series classification archive. Over time it has been incrementally updated, culminating in its current state, HIVE-COTE 1.0. During this time a number of algorithms have been proposed which match the accuracy of HIVE-COTE. We propose comprehensive changes to the HIVE-COTE algorithm which significantly improve its accuracy and usability, presenting this upgrade as HIVE-COTE 2.0. We introduce two novel classifiers, the Temporal Dictionary Ensemble and Diverse Representation Canonical Interval Forest, which replace existing ensemble members. Additionally, we introduce the Arsenal, an ensemble of ROCKET classifiers as a new HIVE-COTE 2.0 constituent. We demonstrate that HIVE-COTE 2.0 is significantly more accurate on average than the current state of the art on 112 univariate UCR archive datasets and 26 multivariate UEA archive datasets.},
url = {https://ueaeprints.uea.ac.uk/id/eprint/80910/},
keywords = {hive-cote,heterogeneous ensembles,multivariate time series,time series classification,software,artificial intelligence ,/dk/atira/pure/subjectarea/asjc/1700/1712}
}
Downloads: 0
{"_id":"XQBp3f7Xi7J5CbWij","bibbaseid":"middlehurst-large-flynn-lines-bostrom-bagnall-hivecote20anewmetaensemblefortimeseriesclassification-2021","author_short":["Middlehurst, M.","Large, J.","Flynn, M.","Lines, J.","Bostrom, A.","Bagnall, A."],"bibdata":{"bibtype":"article","type":"article","note":"Funding Information: This work is supported by the UK Engineering and Physical Sciences Research Council (EPSRC) through an iCASE award sponsored by British Telecom (T206188) and an equipment Grant (T024593). The experiments were carried out on the High Performance Computing Cluster supported by the Research and Specialist Computing Support service at the University of East Anglia and using a Titan X Pascal donated by the NVIDIA Corporation.","month":"December","author":[{"firstnames":["Matthew"],"propositions":[],"lastnames":["Middlehurst"],"suffixes":[]},{"firstnames":["James"],"propositions":[],"lastnames":["Large"],"suffixes":[]},{"firstnames":["Michael"],"propositions":[],"lastnames":["Flynn"],"suffixes":[]},{"firstnames":["Jason"],"propositions":[],"lastnames":["Lines"],"suffixes":[]},{"firstnames":["Aaron"],"propositions":[],"lastnames":["Bostrom"],"suffixes":[]},{"firstnames":["Anthony"],"propositions":[],"lastnames":["Bagnall"],"suffixes":[]}],"title":"HIVE-COTE 2.0: a new meta ensemble for time series classification","journal":"Machine Learning","year":"2021","doi":"10.1007/s10994-021-06057-9","volume":"110","pages":"3211?3243","abstract":"The Hierarchical Vote Collective of Transformation-based Ensembles (HIVE-COTE) is a heterogeneous meta ensemble for time series classification. HIVE-COTE forms its ensemble from classifiers of multiple domains, including phase-independent shapelets, bag-of-words based dictionaries and phase-dependent intervals. Since it was first proposed in 2016, the algorithm has remained state of the art for accuracy on the UCR time series classification archive. Over time it has been incrementally updated, culminating in its current state, HIVE-COTE 1.0. During this time a number of algorithms have been proposed which match the accuracy of HIVE-COTE. We propose comprehensive changes to the HIVE-COTE algorithm which significantly improve its accuracy and usability, presenting this upgrade as HIVE-COTE 2.0. We introduce two novel classifiers, the Temporal Dictionary Ensemble and Diverse Representation Canonical Interval Forest, which replace existing ensemble members. Additionally, we introduce the Arsenal, an ensemble of ROCKET classifiers as a new HIVE-COTE 2.0 constituent. We demonstrate that HIVE-COTE 2.0 is significantly more accurate on average than the current state of the art on 112 univariate UCR archive datasets and 26 multivariate UEA archive datasets.","url":"https://ueaeprints.uea.ac.uk/id/eprint/80910/","keywords":"hive-cote,heterogeneous ensembles,multivariate time series,time series classification,software,artificial intelligence ,/dk/atira/pure/subjectarea/asjc/1700/1712","bibtex":"@article{uea80910,\n note = {Funding Information: This work is supported by the UK Engineering and Physical Sciences Research Council (EPSRC) through an iCASE award sponsored by British Telecom (T206188) and an equipment Grant (T024593). The experiments were carried out on the High Performance Computing Cluster supported by the Research and Specialist Computing Support service at the University of East Anglia and using a Titan X Pascal donated by the NVIDIA Corporation.},\n month = {December},\n author = {Matthew Middlehurst and James Large and Michael Flynn and Jason Lines and Aaron Bostrom and Anthony Bagnall},\n title = {HIVE-COTE 2.0: a new meta ensemble for time series classification},\n journal = {Machine Learning},\n year = {2021},\n doi = {10.1007/s10994-021-06057-9},\n volume = {110},\n pages = {3211?3243},\n abstract = {The Hierarchical Vote Collective of Transformation-based Ensembles (HIVE-COTE) is a heterogeneous meta ensemble for time series classification. HIVE-COTE forms its ensemble from classifiers of multiple domains, including phase-independent shapelets, bag-of-words based dictionaries and phase-dependent intervals. Since it was first proposed in 2016, the algorithm has remained state of the art for accuracy on the UCR time series classification archive. Over time it has been incrementally updated, culminating in its current state, HIVE-COTE 1.0. During this time a number of algorithms have been proposed which match the accuracy of HIVE-COTE. We propose comprehensive changes to the HIVE-COTE algorithm which significantly improve its accuracy and usability, presenting this upgrade as HIVE-COTE 2.0. We introduce two novel classifiers, the Temporal Dictionary Ensemble and Diverse Representation Canonical Interval Forest, which replace existing ensemble members. Additionally, we introduce the Arsenal, an ensemble of ROCKET classifiers as a new HIVE-COTE 2.0 constituent. We demonstrate that HIVE-COTE 2.0 is significantly more accurate on average than the current state of the art on 112 univariate UCR archive datasets and 26 multivariate UEA archive datasets.},\n url = {https://ueaeprints.uea.ac.uk/id/eprint/80910/},\n keywords = {hive-cote,heterogeneous ensembles,multivariate time series,time series classification,software,artificial intelligence ,/dk/atira/pure/subjectarea/asjc/1700/1712}\n}\n\n","author_short":["Middlehurst, M.","Large, J.","Flynn, M.","Lines, J.","Bostrom, A.","Bagnall, A."],"key":"uea80910","id":"uea80910","bibbaseid":"middlehurst-large-flynn-lines-bostrom-bagnall-hivecote20anewmetaensemblefortimeseriesclassification-2021","role":"author","urls":{"Paper":"https://ueaeprints.uea.ac.uk/id/eprint/80910/"},"keyword":["hive-cote","heterogeneous ensembles","multivariate time series","time series classification","software","artificial intelligence","/dk/atira/pure/subjectarea/asjc/1700/1712"],"metadata":{"authorlinks":{}}},"bibtype":"article","biburl":"https://ueaeprints.uea.ac.uk/cgi/search/archive/advanced/export_uea_BibTeX.bib?screen=Search&dataset=archive&_action_export=1&output=BibTeX&exp=0%7C1%7C-date%2Fcreators_name%2Ftitle%7Carchive%7C-%7Ccreators_search_name%3Acreators_search_name%3AANY%3AEQ%3Afinlayson+harvey+mackiewcz+iglesia+taylor+wang+bostrom+zebin+fearne%7Cdate%3Adate%3AALL%3AEQ%3A2020-%7Cdivisions%3Adivisions%3AANY%3AEQ%3ACMP+NBS%7C-%7Ceprint_status%3Aeprint_status%3AANY%3AEQ%3Aarchive%7Cmetadata_visibility%3Ametadata_visibility%3AANY%3AEQ%3Ashow&n=&cache=8648553","dataSources":["PcWqdiiE9bSSvnEiD"],"keywords":["hive-cote","heterogeneous ensembles","multivariate time series","time series classification","software","artificial intelligence","/dk/atira/pure/subjectarea/asjc/1700/1712"],"search_terms":["hive","cote","new","meta","ensemble","time","series","classification","middlehurst","large","flynn","lines","bostrom","bagnall"],"title":"HIVE-COTE 2.0: a new meta ensemble for time series classification","year":2021}