The UCR time series archive

Dau, Son Hoang; Bagnall, Anthony; Kamgar, Kaveh; Yeh, Chin‐Chia Michael; Zhu, Yada; Gharghabi, Shaghayegh; Ratanamahatana, Chotirat Ann; Keogh, Eamonn

doi:10.1109/jas.2019.1911747

Cited by 715 publications

(477 citation statements)

References 36 publications

Supporting

Mentioning

428

Contrasting

Order By: Relevance

“…The vast majority of papers on time series classification consider some improvements on some of the UCR datasets as a necessary and sufficient condition for publication [1] [16]. However, as [22] recently noted, many of the datasets in the archive were created by using a shape-based measure to extract exemplars from a longer time series. Thus, it is tautological to find that shapebased measures work well on this ubiquitous benchmark.…”

Section: 1mentioning

confidence: 99%

Features or Shape? Tackling the False Dichotomy of Time Series Classification

Alaee

Abdoli

Shelton

et al. 2020

Proceedings of the 2020 SIAM International Conference on Data Mining

Self Cite

View full text Add to dashboard Cite

Time series classification is an important task in its own right, and it is often a precursor to further downstream analytics. To date, virtually all works in the literature have used either shape-based classification using a distance measure or feature-based classification after finding some suitable features for the domain. It seems to be underappreciated that in many datasets it is the case that some classes are best discriminated with features, while others are best discriminated with shape. Thus, making the shape vs. feature choice will condemn us to poor results, at least for some classes. In this work, we propose a new model for classifying time series that allows the use of both shape and feature-based measures, when warranted. Our algorithm automatically decides which approach is best for which class, and at query time chooses which classifier to trust the most. We evaluate our idea on real world datasets and demonstrate that our ideas produce statistically significant improvement in classification accuracy.

show abstract

Section: 1mentioning

confidence: 99%

Features or Shape? Tackling the False Dichotomy of Time Series Classification

Alaee

Abdoli

Shelton

et al. 2020

Proceedings of the 2020 SIAM International Conference on Data Mining

Self Cite

View full text Add to dashboard Cite

show abstract

“…The dataset contains 10,934 train cases and is sampled at 2kHz for a series length of 4000. This exceeds the largest train set from ElectricDevices of 8926 and series length from HouseTwenty of 3000 in the 128 UCR datasets [7] (we will donate this data to the archive for the next release). A combination of these factors causes a problem for many standard approaches which would find it difficult to build a model on this data with reasonable time and memory.…”

Section: Whale Acoustics Use Casementioning

confidence: 99%

“…The UCR TSC repository has recently been expanded [7] from 85 to 128 datasets available and including more complex problems such as series of differing length and series with missing values. We make use of this expanded repository in our experiments but exclude datasets with missing values and variable length series due to the currently inability of included classifiers to handle such data.…”

Section: Introductionmentioning

confidence: 99%

Scalable Dictionary Classifiers for Time Series Classification

Middlehurst

Vickers

Bagnall

2019

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

Dictionary based classifiers are a family of algorithms for time series classification (TSC), that focus on capturing the frequency of pattern occurrences in a time series. The ensemble based Bag of Symbolic Fourier Approximation Symbols (BOSS) was found to be a top performing TSC algorithm in a recent evaluation, as well as the best performing dictionary based classifier. A recent addition to the category, the Word Extraction for Time Series Classification (WEASEL), claims an improvement on this performance. Both of these algorithms however have nontrivial scalability issues, taking a considerable amount of build time and space on larger datasets. We evaluate changes to the way BOSS chooses classifiers for its ensemble, replacing its parameter search with random selection. This change allows for the easy implementation of contracting, setting a build time limit for the classifier and check-pointing, saving progress during the classifiers build. To differentiate between the two BOSS ensemble methods we refer to our randomised version as RBOSS. Additionally we test the application of common ensembling techniques to help retain accuracy from the loss of the BOSS parameter search. We achieve a significant reduction in build time without a significant change in accuracy on average when compared to BOSS by creating a size n weighted ensemble selecting the best performers from k randomly chosen parameter sets. Our experiments are conducted on datasets from the recently expanded UCR time series archive. We demonstrate the usability improvements to RBOSS with a case study using a large whale acoustics dataset for which BOSS proved infeasible.

show abstract

“…The archive got an update in 2018, expanding it from 85 to 128 datasets. The new datasets on average have a higher number of training samples and also contain sets with variable length time series to represent many real-world problems [6], ranging from the movement of insect wings to the energy consumption profile of electronic devices. Most datasets in the archive have already been Z-score normalized.…”

Section: Datamentioning

confidence: 99%

Timage – A Robust Time Series Classification Pipeline

Wenninger

Bayerl

Schmidt

et al. 2019

Artificial Neural Networks and Machine Learning – ICANN 2019: Text and Time Series

View full text Add to dashboard Cite

Time series are series of values ordered by time. This kind of data can be found in many real world settings. Classifying time series is a difficult task and an active area of research. This paper investigates the use of transfer learning in Deep Neural Networks and a 2D representation of time series known as Recurrence Plots. In order to utilize the research done in the area of image classification, where Deep Neural Networks have achieved very good results, we use a Residual Neural Networks architecture known as ResNet. As preprocessing of time series is a major part of every time series classification pipeline, the method proposed simplifies this step and requires only few parameters. For the first time we propose a method for multi time series classification: Training a single network to classify all datasets in the archive with one network. We are among the first to evaluate the method on the latest 2018 release of the UCR archive, a well established time series classification benchmarking dataset.

show abstract

The UCR time series archive

Cited by 715 publications

References 36 publications

Features or Shape? Tackling the False Dichotomy of Time Series Classification

Features or Shape? Tackling the False Dichotomy of Time Series Classification

Scalable Dictionary Classifiers for Time Series Classification

Timage – A Robust Time Series Classification Pipeline

Contact Info

Product

Resources

About