SCALE-BOSS-MR: Scalable Time Series Classification Using Multiple Symbolic Representations
Source abstract
Abstract
Time-Series-Classification (TSC) is an important machine learning task for many branches of science. Symbolic representations of time series, especially Symbolic Fourier Approximation (SFA), have been proven very effective for this task, given their abilities to reduce noise. In this paper, we improve upon SCALE-BOSS using multiple symbolic representations of time series. More specifically, the proposed SCALE-BOSS-MR incorporates into the process a variety of window sizes combined with multiple dilation parameters applied to the original and to first-order differences’ time series, with the latter modeling trend information. SCALE-BOSS-MR has been evaluated using the eight datasets with the largest training size of the UCR time series repository. The results indicate that SCALE-BOSS-MR can be instantiated to classifiers that are able to achieve state-of-the-art accuracy and can be tuned for scalability.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Time Series Analysis and Forecasting · Music and Audio Processing · Complex Systems and Time Series Analysis
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Department of Digital Systems, University of Piraeus, 185 34 Piraeus, Greece
References (32)
Show source-provided references
1. Electroencephalogram (EEG) time series classification: Applications in epilepsy10.1007/s10479-006-0076-x ↗
2. Applications of shapelet transform to time series classification of earthquake, wind and wave data10.1016/j.engstruct.2020.111564 ↗
3. Classifying insects on the fly10.1016/j.ecoinf.2013.11.005 ↗
4. Arghandeh, R., and Zhou, Y. (2018). Big Data Application in Power Systems, Elsevier.
5. Scalable time series classification10.1007/s10618-015-0441-y ↗
6. TS-CHIEF: A scalable and accurate forest algorithm for time series classification10.1007/s10618-020-00679-8 ↗
7. ROCKET: Exceptionally fast and accurate time series classification using random convolutional kernels10.1007/s10618-020-00701-z ↗
8. Dempster, A., Schmidt, D.F., and Webb, G.I. (2021, January 14–18). Minirocket: A very fast (almost) deterministic transform for time series classification. Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, Singapore.10.1145/3447548.3467231 ↗
9. Glenis, A., and Vouros, G.A. (2022, January 7–9). SCALE-BOSS: A framework for scalable time-series classification using symbolic representations. Proceedings of the 12th Hellenic Conference on Artificial Intelligence, Corfu, Greece.10.1145/3549737.3549761 ↗
10. MultiRocket: Multiple pooling operators and transformations for fast and effective time series classification10.1007/s10618-022-00844-1 ↗
11. Schäfer, P., and Leser, U. (2023). WEASEL 2.0—A Random Dilated Dictionary Transform for Fast, Accurate and Memory Constrained Time Series Classification. arXiv.10.1007/s10994-023-06395-w ↗
12. Senin, P., and Malinchik, S. (2013, January 7–10). Sax-vsm: Interpretable time series classification using sax and vector space model. Proceedings of the 2013 IEEE 13th International Conference on Data Mining, Dallas, TX, USA.10.1109/ICDM.2013.52 ↗
13. The BOSS is concerned with time series classification in the presence of noise10.1007/s10618-014-0377-7 ↗
14. Schäfer, P., and Högqvist, M. (2012, January 27–30). SFA: A symbolic fourier approximation and index for similarity search in high dimensional datasets. Proceedings of the 15th International Conference on Extending Database Technology, Berlin, Germany.
15. Glenis, A., and Vouros, G.A. (2020, January 19–21). Balancing between scalability and accuracy in time-series classification for stream and batch settings. Proceedings of the International Conference on Discovery Science, Thessaloniki, Greece.10.1007/978-3-030-61527-7_18 ↗
16. Middlehurst, M., Vickers, W., and Bagnall, A. (2019, January 14–16). Scalable dictionary classifiers for time series classification. Proceedings of the International Conference on Intelligent Data Engineering and Automated Learning, Manchester, UK.10.1007/978-3-030-33607-3_2 ↗
17. On time series classification with dictionary-based classifiers10.3233/IDA-184333 ↗
18. Schäfer, P., and Leser, U. (2017, January 6–10). Fast and accurate time series classification with weasel. Proceedings of the 2017 ACM on Conference on Information and Knowledge Management, Singapore.10.1145/3132847.3132980 ↗
19. Nguyen, T.L., and Ifrim, G. (2021). MrSQM: Fast time series classification with symbolic representations. arXiv.
20. Nguyen, T.L., and Ifrim, G. (2022, January 19–23). Fast time series classification with random symbolic subsequences. Proceedings of the International Workshop on Advanced Analytics and Learning on Temporal Data, Grenoble, France.10.1007/978-3-031-24378-3_4 ↗
21. Middlehurst, M., Large, J., Cawley, G., and Bagnall, A. (2020, January 14–18). The temporal dictionary ensemble (TDE) classifier for time series classification. Proceedings of the Joint European Conference on Machine Learning and Knowledge Discovery in Databases, Ghent, Belgium.10.1007/978-3-030-67658-2_38 ↗
22. Bostrom, A., and Bagnall, A. (2017). Transactions on Large-Scale Data-and Knowledge-Centered Systems XXXII, Springer.
23. Proximity forest: An effective and scalable distance-based classifier for time series10.1007/s10618-019-00617-3 ↗
24. Mahato, V., O’Reilly, M., and Cunningham, P. (2018, January 6–7). A Comparison of k-NN Methods for Time Series Classification and Regression. Proceedings of the AICS, Dublin, Ireland.
25. Ester, M., Kriegel, H.P., Sander, J., and Xu, X. (1996, January 2–4). A density-based algorithm for discovering clusters in large spatial databases with noise. Proceedings of the kdd, Portland, OR, USA.
26. DBSCAN revisited, revisited: Why and how you should (still) use DBSCAN10.1145/3068335 ↗
27. BIRCH: An efficient data clustering method for very large databases10.1145/235968.233324 ↗
28. Sculley, D. (2010, January 26–30). Web-scale k-means clustering. Proceedings of the 19th International Conference on World Wide Web, Raleigh, NC, USA.10.1145/1772690.1772862 ↗
29. A simple and fast algorithm for K-medoids clustering10.1016/j.eswa.2008.01.039 ↗
30. pyts: A Python Package for Time Series Classification
31. Scikit-learn: Machine learning in Python
32. Apache spark: A unified engine for big data processing10.1145/2934664 ↗