Extreme Events Characterization on Time Series

  • Marcos Wander Rodrigues Pontifical Catholic University of Minas Gerais
  • Luis Enrique Zárate Pontifical Catholic University of Minas Gerais


The use of sensors in environments where they require constant monitoring has been increasing in recent years. The main goal is to guarantee the effectiveness, safety, and smooth functioning of the system. To identify the occurrence of abnormal events, we propose a methodology that aims to detect patterns that can lead to abrupt changes in the behavior of the sensor signals. To achieve this objective, we provide a strategy to characterize the time series, and we use a clustering technique to analyze the temporal evolution of the sensor system. To validate our methodology, we propose the clusters’ stability index by windowing. Also, we have developed a parameterizable time series generator, which allows us to represent different operational scenarios for a sensor system where extreme anomalies may arise.

Palavras-chave: Anomaly Detection, Characterization of Time Series, Cluster Analysis, Extreme Events, Time Series Analysis


Abdullah, M. Time series motif discovery: dimensions and applications. WIREs Data Mining Knowl Discov vol. 4, pp. 152–159, 2014.

Caires, S. Extreme value analysis: wave data. Tech. Rep. 57, JCOMM Technical Report 57, 2011.

Chavez-Demoulin, V. and Davison, A. C. Modelling the time series extremes. Revstat Statistical Journal vol. 10, pp. 109–133, 03, 2012.

Hotelling, H. The generalization of student’s ratio. In Breakthroughs in Statistics: Foundations and Basic Theory, S. Kotz and N. L. Johnson (Eds.). Springer New York, New York, NY, pp. 54–65, 1992.

Huang, D., Koh, Y. S., and Dobbie, G. Rare pattern mining on data streams. In Proceedings of the 14th International Conference on Data Warehousing and Knowledge Discovery (DaWaK). Springer Berlin Heidelberg, Berlin, Heidelberg, pp. 303–314, 2012.

Lee, T., Zhang, R., Xiao, Y., and Dean, J. Feature extraction methods for time series data in sas enterprise miner. Tech. rep., SAS Institute, 2014.

Nakkeeran, K., Garla, S., and Chakraborty, G. Application of time series clustering using sas enterprise miner for a retail chain. Tech. rep., SAS Global Forum 2012. 04, 2012.

Ranjan, C., Reddy, M., Mustonen, M., Paynabar, K., and Pourak, K. Dataset: Rare event classification in multivariate time series. ArXiv vol. abs/1809.10717, pp. 1–7, 2018.

Serra, A. P. and Zárate, L. E. Characterization of time series for analyzing of the evolution of time series clusters. Expert Systems with Applications 42 (1): 596 – 611, 2015.

Taleb, N. N. The Black Swan: The Impact of the Highly Improbable. Incerto. Random House Publishing Group, London, 2007.

Trovero, M. A. and Leonard, M. J. Time series feature extraction. Tech. rep., SAS 2020-2018, 2018.

Winters, P. R. Forecasting sales by exponentially weighted moving averages. Manag. Science 6 (3): 324–342, 1960.

Yanfei Kang, Rob J Hyndman, F. L. Efficient generation of timeseries with diverse and controllable characteristics. Tech. Rep. 15/18, Monash Business School, 2018.
RODRIGUES, Marcos Wander; ZÁRATE, Luis Enrique. Extreme Events Characterization on Time Series. In: SYMPOSIUM ON KNOWLEDGE DISCOVERY, MINING AND LEARNING (KDMILE), 8. , 2020, Evento Online. Anais [...]. Porto Alegre: Sociedade Brasileira de Computação, 2020 . p. 57-64. ISSN 2763-8944. DOI: https://doi.org/10.5753/kdmile.2020.11959.