Filtering in Preprocessing for Anomaly Detection in Time Series

• Ricardo Fernandes de Araujo Centro Federal de Educação Tecnológica Celso Suckow da Fonseca (CEFET/RJ)
• Laís Baroni Centro Federal de Educação Tecnológica Celso Suckow da Fonseca (CEFET/RJ)
• Eduardo Bezerra Centro Federal de Educação Tecnológica Celso Suckow da Fonseca (CEFET/RJ)
• Gustavo Guedes Centro Federal de Educação Tecnológica Celso Suckow da Fonseca (CEFET/RJ) https://orcid.org/0000-0001-8593-1506
• Esther Pacitti Inria, University of Montpellier, CNRS, LIRMM
• Larrisa Pinheiro Universidade Federal do Rio de Janeiro (UFRJ)
• Diego Salles Centro Federal de Educação Tecnológica Celso Suckow da Fonseca (CEFET/RJ)
• Fabio Porto Laboratório Nacional de Computação Científica (LNCC)
• Eduardo Ogasawara Centro Federal de Educação Tecnológica Celso Suckow da Fonseca (CEFET/RJ)

DOI: https://doi.org/10.5753/sbbd.2026.249344

Resumo

Filtering is often adopted as a preprocessing step in time series pipelines, although its impact on anomaly detection remains unclear. We evaluate 14 filters with an ARIMA-based detector on the GECCO Industrial Challenge 2018 and Yahoo! Webscope S5 benchmarks. Filtering improves F1-Score in only 25.36% of the runs; in the remaining 74.64%, it either fails to improve performance or degrades it. The results reveal a trade-off between stability and potential gains. Adaptive filters are more consistent, whereas decomposition-based filters can yield larger gains but also more frequent degradation. Within this ARIMA-based protocol, filtering does not behave uniformly across series and should be treated as a context-dependent engineering choice.

Palavras-chave: Anomaly Detection, Time Series Analysis, Data Preprocessing, Temporal Filtering, ARIMA Models

Referências

Contreras, J., Espínola, R., Nogales, F. J., and Conejo, A. J. (2003). ARIMA models to predict next-day electricity prices. IEEE Transactions on Power Systems, 18:1014 – 1020.

Gea, C., Lima, J., Bezerra, E., and Ogasawara, E. (2021). Análise de métodos de tratamento de outliers para predição dos retornos de índices de ações negociados em bolsa. In Anais do Simpósio Brasileiro de Banco de Dados (SBBD), pages 277–282. SBC.

Gupta, M., Gao, J., Aggarwal, C. C., and Han, J. (2014). Outlier Detection for Temporal Data: A Survey. IEEE Transactions on Knowledge and Data Engineering, 26:2250 – 2267.

Hyndman, R. J. and Athanasopoulos, G. (2018). Forecasting: principles and practice. OTexts.

Inoue, J., Yamagata, Y., Chen, Y., Poskitt, C. M., and Sun, J. (2017). Anomaly detection for a water treatment system using unsupervised machine learning. In R, G., G, K., V, R., X, W., L, M., S, A., X, N., and G, D., editors, IEEE International Conference on Data Mining Workshops, ICDMW, volume 2017-November, pages 1058 – 1065. IEEE Computer Society.

Ishimtsev, V., Bernstein, A., Burnaev, E., and Nazarov, I. (2017). Conformal k-NN Anomaly Detector for Univariate Data Streams. In Proceedings of the Sixth Workshop on Conformal and Probabilistic Prediction and Applications, volume 60 of Proceedings of Machine Learning Research, pages 213–227. PMLR.

Mao, Y., Qi, H., Chen, X., and Li, X. (2017). Event Detection with Multivariate Water Parameters in the Water Monitoring Applications. In Proceedings - 4th IEEE International Conference on Cyber Security and Cloud Computing, CSCloud 2017 and 3rd IEEE International Conference of Scalable and Smart Cloud, SSC 2017, pages 321 – 326.

Ogasawara, E., Alexandrino, F., Gea, C., Santos, D., Salles, R., Birindiba, V., Pacheco, C., Bezerra, E., Pacitti, E., Porto, F., and Carvalho, D. (2025a). tspredit: Time Series Prediction with Integrated Tuning.

Ogasawara, E., Salles, R., Porto, F., and Pacitti, E. (2025b). Event Detection in Time Series. Synthesis Lectures on Data Management. Springer Nature Switzerland, Cham, 1 edition.

Pena, E. H. M., De Assis, M. V. O., and Proença, M. L. (2013). Anomaly Detection Using Forecasting Methods ARIMA and HWDS. In Proceedings - International Conference of the Chilean Computer Science Society, SCCC, volume 0, pages 63 – 66.

Salles, R., Escobar, L., Baroni, L., Zorrilla, R., Ziviani, A., Kreischer, V., Delicato, F., Pires, P. F., Maia, L., Coutinho, R., Assis, L., and Ogasawara, E. (2020). Harbinger: Um framework para integração e análise de métodos de detecção de eventos em séries temporais. In Anais do Simpósio Brasileiro de Banco de Dados (SBBD), pages 73–84. SBC.

Yaacob, A. H., Tan, I. K., Chien, S. F., and Tan, H. K. (2010). ARIMA based network anomaly detection. In 2nd International Conference on Communication Software and Networks, ICCSN 2010, pages 205 – 209.

Zamanzadeh Darban, Z., Webb, G. I., Pan, S., Aggarwal, C., and Salehi, M. (2024). Deep Learning for Time Series Anomaly Detection: A Survey. ACM Computing Surveys, 57:1–42.

Zhu, B. and Sastry, S. (2011). Revisit dynamic ARIMA based anomaly detection. In Proceedings - 2011 IEEE International Conference on Privacy, Security, Risk and Trust and IEEE International Conference on Social Computing, PASSAT/SocialCom 2011, pages 1263 – 1268.