Generation of Spatial Time Series Data
Alena Mikhailovna KROPACHEVA,
Dmitry Viktorovich GIRDYUK,
Illarion Lavrentievich IOV,
Anton Yurievich PERSHIN https://doi.org/10.15514/ISPRAS-2024-36(4)-11
Abstract
In the era of deep learning, global-local deep neural networks are gradually replacing statistical approaches for time-series forecasting, especially for the spatiotemporal modeling field. However, the development of such methods is hindered by the lack of open benchmark datasets in this research domain. Generating synthetic data is an alternative solution to data collection, but prior works focus mainly on generating uncorrelated independent time series. In this work, we present a method for spatially correlated time-series generation. It uses a set of parametric autoregressive models for univariate time series generation in combination with the approach for sampling model parameters which allows one to simulate spatial relationships. We describe its implementation and conduct experiments showing the validity of the data for spatiotemporal modeling.
Supporting agencies: The research was carried out within the state assignment of Ministry of Science and Higher Education of the Russian Federation (project No. FSER-2024-0004).
About the Authors
Alena Mikhailovna KROPACHEVA
St. Petersburg State University
Russian Federation
Russian Federation
Undergraduate student at the Faculty of Mathematics and Mechanics of Saint Petersburg State University. Her research interests include time series analysis, classical computer vision algorithms, machine learning, and deep learning.
Dmitry Viktorovich GIRDYUK
St. Petersburg State University
Russian Federation
Russian Federation
Assistant of the Department of Functional Systems Diagnostics of Saint Petersburg State University. His research interests include computer vision, time series analysis, and mathematical immunology.
Illarion Lavrentievich IOV
ITMO University
Russian Federation
Russian Federation
Post graduate student at Natural Systems Simulation Lab at ITMO University. Research interests: automated machine learning, large language models applications to optimization, time series analysis.
Anton Yurievich PERSHIN
St. Petersburg State University
Russian Federation
Russian Federation
Ph.D., associate professor of the Department of Fundamental Informatics and Distributed Systems of Saint Petersburg State University. Research interests: chaotic dynamical systems, transition to turbulence, numerical methods for stability analysis, time series analysis.
References
1. Y. Hahn, T. Langer, R. Meyes, and T. Meisen, “Time series dataset survey for forecasting with deep learning,” Forecasting, vol. 5, no. 1, pp. 315–335, 2023, Available at: https://www.mdpi.com/25719394/5/1/17, accessed 02.04.2024.
2. S. Makridakis, E. Spiliotis, and V. Assimakopoulos, “The m4 competition: 100,000 time series and 61 forecasting methods,” International Journal of Forecasting, vol. 36, no. 1, pp. 54–74, 2020, m4 Competition.
3. S. Makridakis, E. Spiliotis, and V. Assimakopoulos, “The m5 competition: Background, organization, and implementation,” International Journal of Forecasting, vol. 38, no. 4, pp. 1325–1336, 2022, special Issue: M5 competition.
4. M. Abolghasemi, G. Tarr, and C. Bergmeir, “Machine learning applications in hierarchical time series forecasting: Investigating the impact of promotions,” International Journal of Forecasting, vol. 40, no. 2, pp. 597–615, 2024.
5. Y. Li, R. Yu, C. Shahabi, and Y. Liu, “Graph convolutional recurrent neural network: Data-driven traffic forecasting,” ArXiv, vol. abs/1707.01926, 2017, Available at: https://api.semanticscholar.org/CorpusID:195346050, , accessed 02.04.2024.
6. Y. Kang, R. J. Hyndman, and F. Li, “Gratis: Generating time series with diverse and controllable characteristics,” 2019.
7. Q. Wen, T. Zhou, C. Zhang, W. Chen, Z. Ma, J. Yan, and L. Sun, “Transformers in time series: A survey,” arXiv preprint arXiv:2202.07125, 2022.
8. J. A. Miller, M. Aldosari, F. Saeed, N. H. Barna, S. Rana, I. B. Arpinar, and N. Liu, “A survey of deep learning and foundation models for time series forecasting,” arXiv preprint arXiv:2401.13912, 2024.
9. S. Wang, Y. Du, X. Guo, B. Pan, Z. Qin, and L. Zhao, “Controllable data generation by deep learning: A review,” ACM Comput. Surv., mar 2024, Available at: https://doi.org/10.1145/3648609, accessed 02.04.2024.
10. “timeseries-generator”, Available at: https://github.com/NikeInc/timeseries-generator, accessed 18.05.2024.
11. “mockseries”, Available at: https://github.com/cyrilou242/mockseries, accessed 18.05.2024.
12. M. Dogariu, L.-D. S¸tefan, B. A. Boteanu, C. Lamba, B. Kim, and B. Ionescu, “Generation of realistic synthetic financial time-series,” ACM Transactions on Multimedia Computing, Communications, and Applications, vol. 18, no. 4, p. 1–27, 2022.
13. Y. Li, R. Yu, C. Shahabi, and Y. Liu, “Diffusion convolutional recurrent neural network: Data-driven traffic forecasting,” arXiv preprint arXiv:1707.01926, 2017.
14. P. J. Brockwell and R. A. Davis, Introduction to time series and forecasting. Springer, 2016.
15. P. W. Talbot, C. Rabiti, A. Alfonsi, C. Krome, M. R. Kunz, A. Epiney, C. Wang, and D. Mandelli, “Correlated synthetic time series generation for energy system simulations using fourier and arma signal processing,” International Journal of Energy Research, vol. 44, no. 10, p. 8144–8155, 2020.
16. “mockseries documentation.”, Available at: https://mockseries.catheu.tech/docs/tutorials/interaction-scalaroperations, accessed 19.05.2024.
17. К.В. Воронцов, “Машинное обучение. Прогнозирование временных рядов. К.В. Воронцов, Школа анализа данных, Яндекс”, 2020, Available at: https://youtu.be/Rmh6b96u6UU?si=o3I20WlIP5EKW2kw (in Russian), accessed 08.10.2023.
18. R. J. Hyndman and G. Athanasopoulos, Forecasting: Principles and practice, 3rd ed. OTexts, 2021.
19. “Python”, Available at: https://docs.python.org/3/tutorial/index.html, accessed 11.04.2024.
20. “Numpy”, Available at: https://numpy.org/doc/stable/, accessed 19.05.2024.
21. “matplotlib”, Available at: https://matplotlib.org/stable/, accessed 17.05.2024.
22. “scikit-learn”, Available at: https://scikitlearn.org/stable/index.html, accessed 19.05.2024.
23. “time-series-generator”, Available at: https://github.com/stilmmo/time-series-generator, accessed 01.06.2024.
24. Y. Li, R. Yu, C. Shahabi, and Y. Liu, “Diffusion convolutional recurrent neural network: Data-driven traffic forecasting,” in International Conference on Learning Representations (ICLR ’18), 2018.
Review
For citations:
KROPACHEVA A.M., GIRDYUK D.V., IOV I.L., PERSHIN A.Yu. Generation of Spatial Time Series Data. Proceedings of the Institute for System Programming of the RAS (Proceedings of ISP RAS). 2024;36(4):143-154. https://doi.org/10.15514/ISPRAS-2024-36(4)-11
Email this article 