The great time series classification bake off: a review and experimental evaluation of recent algorithmic advances
Reads0
Chats0
TLDR
This work implemented 18 recently proposed algorithms in a common Java framework and compared them against two standard benchmark classifiers (and each other) by performing 100 resampling experiments on each of the 85 datasets, indicating that only nine of these algorithms are significantly more accurate than both benchmarks.Abstract:
In the last 5 years there have been a large number of new time series classification algorithms proposed in the literature. These algorithms have been evaluated on subsets of the 47 data sets in the University of California, Riverside time series classification archive. The archive has recently been expanded to 85 data sets, over half of which have been donated by researchers at the University of East Anglia. Aspects of previous evaluations have made comparisons between algorithms difficult. For example, several different programming languages have been used, experiments involved a single train/test split and some used normalised data whilst others did not. The relaunch of the archive provides a timely opportunity to thoroughly evaluate algorithms on a larger number of datasets. We have implemented 18 recently proposed algorithms in a common Java framework and compared them against two standard benchmark classifiers (and each other) by performing 100 resampling experiments on each of the 85 datasets. We use these results to test several hypotheses relating to whether the algorithms are significantly more accurate than the benchmarks and each other. Our results indicate that only nine of these algorithms are significantly more accurate than both benchmarks and that one classifier, the collective of transformation ensembles, is significantly more accurate than all of the others. All of our experiments and results are reproducible: we release all of our code, results and experimental details and we hope these experiments form the basis for more robust testing of new algorithms in the future.read more
Citations
More filters
Posted Content
Matrix Profile XXII: Exact Discovery of Time Series Motifs under DTW
TL;DR: This work presents the first scalable exact method to discover time series motifs under Dynamic Time Warping, and automatically performs the best trade-off between time-to-compute and tightness-of-lower-bounds for a novel hierarchy of lower bounds representation it introduces.
Journal ArticleDOI
Feature extraction from unequal length heterogeneous EHR time series via dynamic time warping and tensor decomposition
TL;DR: A novel approach to tackle the issue of irregularly sampled, unequal length EHR time series using dynamic time warping and tensor decomposition and produces outstanding classification performance in terms of AUROC, AUPRC and accuracy compared with the baseline methods: LSTM and DTW-KNN.
Journal ArticleDOI
Artificial neural networks for classifying the time series sensor data generated by medical detection dogs
TL;DR: To achieve a useful level of accuracy, it was found that the models needed to be trained using only those data samples where the dog had correctly classified the scent sample, and model hyperparameters were tuned to improve accuracy.
Journal ArticleDOI
Can the global modeling technique be used for crop classification
Sylvain Mangiarotti,Amit Kumar Sharma,Samuel Corgne,Laurence Hubert-Moy,Laurent Ruiz,Muddu Sekhar,Yann Kerr +6 more
TL;DR: In this article, a new classification technique based on the global modeling technique is introduced to cope with the difficulties of crop detection from remote sensed images, which is of major interest for land use and land cover mapping.
Knowledge Extraction with Interval Temporal Logic Decision Trees.
TL;DR: This paper introduces Temporal C4.5, that allows the extraction of temporal decision trees from undiscretized multivariate time series, its implementation is described, called Temporal J48, and the outcome of a set of experiments with the latter is discussed, comparing the results with those obtained by other, classical, multivariateTime series classification methods.
References
More filters
Journal ArticleDOI
The WEKA data mining software: an update
TL;DR: This paper provides an introduction to the WEKA workbench, reviews the history of the project, and, in light of the recent 3.6 stable release, briefly discusses what has been added since the last stable version (Weka 3.4) released in 2003.
Journal Article
Statistical Comparisons of Classifiers over Multiple Data Sets
TL;DR: A set of simple, yet safe and robust non-parametric tests for statistical comparisons of classifiers is recommended: the Wilcoxon signed ranks test for comparison of two classifiers and the Friedman test with the corresponding post-hoc tests for comparisons of more classifiers over multiple data sets.
Book ChapterDOI
Domain-adversarial training of neural networks
Yaroslav Ganin,Evgeniya Ustinova,Hana Ajakan,Pascal Germain,Hugo Larochelle,François Laviolette,Mario Marchand,Victor Lempitsky +7 more
TL;DR: In this article, a new representation learning approach for domain adaptation is proposed, in which data at training and test time come from similar but different distributions, and features that cannot discriminate between the training (source) and test (target) domains are used to promote the emergence of features that are discriminative for the main learning task on the source domain.
Journal ArticleDOI
Experiencing SAX: a novel symbolic representation of time series
TL;DR: The utility of the new symbolic representation of time series formed is demonstrated, which allows dimensionality/numerosity reduction, and it also allows distance measures to be defined on the symbolic approach that lower bound corresponding distance measuresdefined on the original series.
Journal ArticleDOI
Querying and mining of time series data: experimental comparison of representations and distance measures
TL;DR: An extensive set of time series experiments are conducted re-implementing 8 different representation methods and 9 similarity measures and their variants and testing their effectiveness on 38 time series data sets from a wide variety of application domains to provide a unified validation of some of the existing achievements.