Contamination Map Methodology

Time series / Dataset

UEA Multivariate Time Series Classification Archive (2018)

First iteration (2018 vintage) of the multivariate time-series classification archive by UEA and UCR: 30 datasets, formatted to equal length, with no missing values and with train/test splits provided (splits given by the data donors were retained). Sources include the UCI Machine Learning archive, Brain Computer Interface competitions, Kaggle competitions and datasets made by the authors. Hosted at www.timeseriesclassification.com.

Also known as
UEA archive, UEA multivariate time series classification archive, UEA MTSC archive, UEA 2018, UEA30, UEA multivariate archive
Domain
mixed (human activity recognition, motion, ECG, EEG/MEG, audio spectra, other)
Test items from
2018-10-31
Evidence
“arXiv:1811.00075v1 [cs.LG] 31 Oct 2018”
arxiv.org · paper · header · retrieved 2026-10-15
“The 2018 vintage consists of 30 datasets with a wide range of cases, dimensions and series lengths.”
arxiv.org · paper · Section 1 · retrieved 2026-10-15
Homepage
www.timeseriesclassification.com

Exposure by model

BenchmarkModelLevelReason
Time-series classification Chronos-2-Synth Not seen absent from the complete list of training data
Time-series classification Chronos-2 Not seen absent from the complete list of training data
Time-series classification Chronos-Bolt Unknown training data not disclosed in enough detail
Time-series classification Chronos (T5) Not seen absent from the complete list of training data
Time-series classification DADA (General Time Series Anomaly Detector with Adaptive Bottlenecks and Dual Adversarial Decoders) Unknown training data not disclosed in enough detail
Time-series classification Granite-TimeSeries-TTM-R1 Not seen absent from the complete list of training data
Time-series classification Granite-TimeSeries-TTM-R2.1 Not seen absent from the complete list of training data
Time-series classification Granite-TimeSeries-TTM-R2 Not seen absent from the complete list of training data
Time-series classification Granite-TTM-R3 Unknown training data not disclosed in enough detail
Time-series classification Lag-Llama Not seen absent from the complete list of training data
Time-series classification Mantis (Mantis-8M) Training data seen only the training split was used, not the benchmark's test data; the test portion was explicitly excluded
Time-series classification Mantis+ (MantisPlus) Not seen absent from the complete list of training data
Time-series classification MantisV2 Not seen absent from the complete list of training data
Time-series classification Moirai-1.0-R Not seen absent from the complete list of training data
Time-series classification Moirai-1.1-R Unknown training data not disclosed in enough detail
Time-series classification Moirai-2.0-R Unknown training data not disclosed in enough detail
Time-series classification Moirai-MoE-1.0-R Not seen absent from the complete list of training data
Time-series classification MOMENT-1 Training data seen only the training split was used, not the benchmark's test data (via Time Series Pile)
Time-series classification NuTime Training data seen only the training split was used, not the benchmark's test data; the test portion was explicitly excluded
Time-series classification Sundial (Timer 3.0) Seen, split unclear the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD))
Time-series classification Time-MoE Unknown training data not disclosed in enough detail
Time-series classification Time-RCD (TimeRCD) Not seen absent from the complete list of training data
Time-series classification Timer-S1 Seen, split unclear the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD))
Time-series classification Timer-XL (timer-base-84m) Seen, split unclear the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD))
Time-series classification Timer Seen, split unclear the dataset was used; the source does not say which part (via Unified Time Series Dataset (UTSD))
Time-series classification TimeRadar Unknown training data not disclosed in enough detail
Time-series classification TimesFM 1.0 Unknown training data not disclosed in enough detail
Time-series classification TimesFM 2.0 Unknown training data not disclosed in enough detail
Time-series classification TimesFM 2.5 Unknown training data not disclosed in enough detail
Time-series classification TimesFM 3.0 Unknown training data not disclosed in enough detail
Time-series classification TiRex 1.1 (gifteval) Unknown training data not disclosed in enough detail
Time-series classification TiRex-2-fevbench (TiRex-2-f) Not seen absent from the complete list of training data
Time-series classification TiRex-2-gifteval-pretrain (TiRex-2-gp) Unknown training data not disclosed in enough detail
Time-series classification TiRex-2-gifteval-zs (TiRex-2-g) Not seen absent from the complete list of training data
Time-series classification TiRex-2 Unknown training data not disclosed in enough detail
Time-series classification TiRex Not seen absent from the complete list of training data
Time-series classification Toto-Open-Base-1.0 Unknown training data not disclosed in enough detail
Time-series classification Toto-2.0-2.5B-FT Unknown training data not disclosed in enough detail
Time-series classification Toto 2.0 Unknown training data not disclosed in enough detail
Time-series classification Toto (2024 technical report) Unknown training data not disclosed in enough detail
Time-series classification TSPulse r1 Not seen the authors state the dataset was excluded from training
Time-series classification TTM-R3 Unknown training data not disclosed in enough detail
Time-series classification Tiny Time Mixer (TTM) Research-Use R2 Not seen absent from the complete list of training data
Time-series classification UniTS few-shot anomaly detection checkpoints Training data seen via base model UniTS: only the training split was used, not the benchmark's test data
Time-series classification UniTS Training data seen only the training split was used, not the benchmark's test data

Notes

published_from = arXiv v1 date of the archive paper. Many constituent datasets were public earlier in other forms (UCI repository, BCI and Kaggle competitions, some also in the Baydogan multivariate archive). No licence stated in the paper.

Sources