The UEA multivariate time series classification archive, 2018

By Anthony J. Bagnall, Hoang Anh Dau, Jason Lines, Michael D. Flynn, James Large, Aaron Bostrom, Paul Southam, Eamonn J. Keogh. arXiv.org, 2018

616 citations recorded by the source corpus.

In 2002, the UCR time series classification archive was first released with sixteen datasets. It gradually expanded, until 2015 when it increased in size from 45 datasets to 85 datasets. In October 2018 more datasets were added, bringing the total to 128. The new archive contains a wide range of problems, including variable length series, but it still only contains univariate time series classification problems. One of the motivations for introducing the archive was to encourage researchers to perform a more rigorous evaluation of newly proposed time series classification (TSC) algorithms. It has worked: most recent research into TSC uses all 85 datasets to evaluate algorithmic advances. Research into multivariate time series classification, where more than one series are associated with each class label, is in a position where univariate TSC research was a decade ago. Algorithms are evaluated using very few datasets and claims of improvement are not based on statistical comparisons. We aim to address this problem by forming the first iteration of the MTSC archive, to be hosted at the website this http URL. Like the univariate archive, this formulation was a collaborative effort between researchers at the University of East Anglia (UEA) and the University of California, Riverside (UCR). The 2018 vintage consists of 30 datasets with a wide range of cases, dimensions and series lengths. For this first iteration of the archive we format all data to be of equal length, include no series with missing data and provide train/test splits.

Subfields: Artificial Intelligence, Signal Processing

View Paper: https://www.semanticscholar.org/paper/d8abb8206b913d185b4bd406880131c13759a6ff

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Signal Processing

Citations in the corpus

Deep Learning in Human Activity Recognition with Wearable Sensors: A Review on Advances (2021, background): “Recently, researchers from the University of East Anglia have collaborated with UCR to generate a new collection of datasets, which includes 9 categories of HAR: BasicMotions, Cricket, Epilepsy, ERing, Handwriting, Libras, NATOPS, RacketSports, and UWaveGestureLibrary [24].”
HIVE-COTE 2.0: a new meta ensemble for time series classification (2021, methodology): “HC2 is signiﬁcantly more accurate than all these algorithms on the UEA MTSC archive (Bagnall et al., 2018).”
TapNet: Multivariate Time Series Classification with Attentional Prototypical Network (2020, methodology): “…of research domains and applications, multivariate time series classification models have been used in many different real-world applications such as Human Activity Recognition (Minnen et al. 2006), EEG/ECG data analysis, (Bagnall et al. 2018)(Wang et al. ∗These two authors contributed equally.”
The great multivariate time series classification bake off: a review and experimental evaluation of recent algorithmic advances (2020, background): “The UEA MTSC archive (Bagnall et al. 2018)was formed to overcome these problems.”
TimesNet: Temporal 2D-Variation Modeling for General Time Series Analysis (2022, methodology): “Concretely, we select 10 multivariate datasets from UEA Time Series Classiﬁcation Archive (Bagnall et al., 2018), covering the gesture, action and audio recognition, medical diagnosis by heartbeat monitoring and other practical tasks.”
Unsupervised Scalable Representation Learning for Multivariate Time Series (2019, methodology): “However, we provide a comparison with DTWD as baseline using results provided by Bagnall et al. (2018).”
Long Expressive Memory for Sequence Modeling (2021, background): “The goal of this task (Bagnall et al., 2018) is to classify worms as belonging to either the wild-type or four different mutants, based on 259 very long sequences (length N = 17984 ) measuring the motion of a worm.”
Neural Controlled Differential Equations for Irregular Time Series (2020, methodology): “To do this, we consider the CharacterTrajectories dataset from the UEA time series classification archive [31].”
Temporal Convolutional Neural Network for the Classification of Satellite Image Time Series (2018, neutral): “In addition, they have been mainly developed for uni-variate time series, even if the community is actively now proposing adaptations of these algorithms for multi-variate datasets [79].”
TS2Vec: Towards Universal Representation of Time Series (2021, methodology): “2019) and UEA archive (Bagnall et al. 2018) are adopted for evaluation.”
Time to Embed: Unlocking Foundation Models for Time Series with Channel Descriptions (2025, unknown): “We expand on our datasets Appendix G, training, and baselines in Appendix I. Classification To evaluate the discriminative power of our learned representations, we benchmark our model on the widely-used UEA multivariate time series classification suite (Bagnall et al., 2018).”