Detección de anomalías en series temporales multivariantes (MTS) basada en métodos de estimación de la densidad.
| dc.contributor.advisor | Gonzalez Osorio, Fabio Augusto | |
| dc.contributor.author | Rodriguez Peraza, Andres Francisco | |
| dc.contributor.researchgroup | Mindlab | |
| dc.date.accessioned | 2026-02-10T12:41:03Z | |
| dc.date.available | 2026-02-10T12:41:03Z | |
| dc.date.issued | 2025 | |
| dc.description | Ilustraciones, gráficos | spa |
| dc.description.abstract | La detección de anomalías en series temporales multivariantes (MTS) es fundamental para la identificación temprana de comportamientos inusuales en datos secuenciales de múltiples variables. Esta trabajo de grado propone un enfoque basado en métodos de estimación de densidad, que permite modelar de forma probabilística la distribución de datos normales y detectar desviaciones significativas como anomalías. Se implementa un modelo de detección de anomalías en series temporales multivariantes basado en métodos de estimación de densidad, y se evalúa exclusivamente sobre el dataset Server Machine Dataset (SMD), el cual proporciona registros reales de múltiples máquinas con diferentes configuraciones operativas. Este conjunto permite validar el modelo en escenarios con alta dimensionalidad, datos desbalanceados y comportamiento variable entre contextos. Los resultados se analizan usando métricas estándar AUC-PR y F1-score. Se diseñó un experimento comparativo frente a métodos del estado del arte, mostrando ventajas en contextos con alta dimensionalidad y datos desbalanceados. Los resultados sugieren que la estimación de densidad permite capturar mejor los patrones normales y mejorar la detección de desviaciones significativas en datos multivariantes. (Texto tomado de la fuente) | spa |
| dc.description.abstract | Anomaly detection in multivariate time series (MTS) is essential for identifying atypical behaviors in complex systems. This thesis proposes an approach based on density estimation methods, which probabilistically model the distribution of normal data to detect significant deviations as anomalies. A multivariate time series anomaly detection model based on density estimation methods is implemented, and it is evaluated exclusively on the Server Machine Dataset (SMD). This dataset contains real-world measurements from multiple machines with diverse operational characteristics, providing a challenging and realistic environment to assess the model’s generalization capabilities. The evaluation focuses on high-dimensional data, class imbalance, and variability across machine types. Results are analyzed using standard metrics such as precision, recall, and F1-score. The proposed approach is expected to outperform traditional models in terms of robustness and generalization, contributing to the advancement of unsupervised techniques for anomaly detection in real-world scenarios. | eng |
| dc.description.degreelevel | Maestría | |
| dc.description.degreename | Magíster en Ingeniería de Sistemas y computación | |
| dc.description.technicalinfo | Este trabajo se desarrolló mediante la implementación computacional de un modelo de detección de anomalías en series temporales multivariantes basado en métodos de estimación de densidad, específicamente el enfoque INQMAD (Incremental Quantum Measurement Anomaly Detection), el cual utiliza representaciones de matrices de densidad y proyecciones en espacios de Hilbert simulados mediante características de Fourier adaptativas. La solución fue implementada en un entorno de programación científica, empleando herramientas orientadas al análisis de datos y aprendizaje automático. Los principales componentes técnicos utilizados fueron: • Lenguaje de programación: Python • Librerías de análisis numérico: NumPy, SciPy • Procesamiento de datos: Pandas • Aprendizaje automático y modelos base comparativos: Scikit-learn • Modelos de aprendizaje profundo (línea base): Implementaciones basadas en redes LSTM, Autoencoders y modelos generativos tipo VAE-GMM • Visualización de datos: Matplotlib y Seaborn | spa |
| dc.format.extent | x, 61 páginas | |
| dc.format.mimetype | application/pdf | |
| dc.identifier.instname | Universidad Nacional de Colombia | spa |
| dc.identifier.reponame | Repositorio Institucional Universidad Nacional de Colombia | spa |
| dc.identifier.repourl | https://repositorio.unal.edu.co/ | spa |
| dc.identifier.uri | https://repositorio.unal.edu.co/handle/unal/89439 | |
| dc.publisher | Universidad Nacional de Colombia | |
| dc.publisher.branch | Universidad Nacional de Colombia - Sede Bogotá | |
| dc.publisher.faculty | Facultad de Ingeniería | |
| dc.publisher.place | Bogotá, Colombia | |
| dc.publisher.program | Bogotá - Ingeniería - Maestría en Ingeniería - Ingeniería de Sistemas y Computación | |
| dc.relation.references | Aytekin,C.;Si,S.: DeepLearning-BasedMultivariateTimeSeriesAnomalyDetection with Non-Linear Dimensionality Reduction. En: IEEE Access 8 (2020), p. 54171–54182 | |
| dc.relation.references | Ch ; ola, Banerjee A. ; Kumar, V.: Anomaly detection: A survey. En: ACM computing surveys (CSUR) 41 (2009), Nr. 3, p. 1–58 | |
| dc.relation.references | Che, Z.; Purushotham, S.; Cho, K.; Sontag, D.; Liu, Y. Recurrent Neural Networks for Multivariate Time Series with Missing Values. En: Scientific Reports 8 (2018), Nr. 1, p. 1–12. | |
| dc.relation.references | Ch ; ola, Banerjee A. ; Kumar, V.: Anomaly detection: A survey. En: ACM computing surveys (CSUR) 41 (2009), Nr. 3, p. 1–58 | |
| dc.relation.references | Chen, J.; Han, J. Anomaly Detection with Multiple-Hierarchical Features and Soft Labeling for Multivariate Time Series. En: Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2019. | |
| dc.relation.references | Chen, T. ; Guestrin, C.: XGBoost: A scalable tree boosting system. En: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2018, p. 785–794 | |
| dc.relation.references | Chen, Xu H. He B. Liu Y. Liu X. ; Wu, D.: Anomaly detection in multivariate time series with uncertainty quantification. En: IEEE Transactions on Knowledge and Data Engineering 33 (2021), Nr. 5, p. 1955–1968 | |
| dc.relation.references | Darban, Webb G. I. Pan S. Aggarwal C. C. ; Salehi, M.: Deep Learning for Time Series Anomaly Detection: A Survey. En: arXiv preprint arXiv:2211.05244 (2022) | |
| dc.relation.references | Filonov, Pavel ; Lavrentyev, Andrey ; Vorontsov, Artem: Multivariate industrial time series with cyber-attack simulation: Fault detection using an lstm-based predictive data model. En: arXiv preprint arXiv:1612.06676 (2016) | |
| dc.relation.references | Gallego-Mejia, Bustos-Brinez O. A. ; González, F. A.: InQMAD: Incremental Quantum Measurement Anomaly Detection. En: 2022 IEEE International Conference on Data Mining Workshops (ICDMW), IEEE, November 2022, p. 787–796 | |
| dc.relation.references | Gallego-Mejia, J. A. ; González, F. A.: DEMANDE: Density Matrix Neural Den- sity Estimation. En: IEEE (2023) | |
| dc.relation.references | Grubbs, F. E.: Procedures for detecting outlying observations in samples. (1969) | |
| dc.relation.references | Guo, Sun Y. ; Wang, L.: Multivariate time series anomaly detection with missing data based on stacked autoencoder. En: IEEE Access 8 (2020), p. 129933–129945 | |
| dc.relation.references | Hassan, Mohammed A. A. ; Al-Betar, M. A.: A Review on Deep Learning-Based Anomaly Detection Techniques for Time-Series Data. En: SN Computer Science 1 (2020), Nr. 4, p. 1–32 | |
| dc.relation.references | He, Yangdong ; Zhao, Jiabao: Temporal convolutional networks for anomaly detection in time series. En: Journal of Physics: Conference Series Vol. 1213 IOP Publishing, 2019, p. 042050 | |
| dc.relation.references | Hido, S. ; Kashima, H.: Anomaly detection in multivariate time series with variational autoencoder. En: Proceedings of the International Conference on Data Mining (DMIN), 2016, p. 11–17 | |
| dc.relation.references | Hochenbaum, Vallis O. S. ; Kejariwal, A.: Automatic detection of anomalies in multivariate time series. En: Journal of Intelligent Information Systems 49 (2017), Nr. 2, p. 287–305 | |
| dc.relation.references | Khaleghi, A. ; Moin, M. S.: Improved anomaly detection in surveillance videos based on a deep learning method. En: 2018 8th Conference of AI Robotics and 10th RoboCup Iranopen International Symposium (IRANOPEN), IEEE, April 2018, p. 73–81 | |
| dc.relation.references | Kiran, B. R. ; Kumar, V.: Ensemble of feature-based clustering and classification for anomaly detection in time-series data. En: IEEE Transactions on Knowledge and Data Engineering 29 (2017), Nr. 9, p. 1907–1920 | |
| dc.relation.references | Kiran,ThomasD.M.;Parakkal,R.: Anoverviewofdeeplearningbasedmethodsfor unsupervised and semi-supervised anomaly detection in videos. En: Journal of Imaging 4 (2018), Nr. 2, p. 36 | |
| dc.relation.references | Kiran R, Singh A.: Multivariate time series anomaly detection: A survey. En: IEEE Access 8 (2020), p. 193047–193080 | |
| dc.relation.references | Kwon, Y. W. ; Cho, S. B.: A hybrid approach for time series anomaly detection using LSTM and a sliding window-based statistical method. En: Applied Sciences 9 (2019), Nr. 1, p. 109 | |
| dc.relation.references | Lai, Y. K. ; Ng, W. W.: Unsupervised Anomaly Detection in Multivariate Time Series Data Using Clustering. En: Neurocomputing 275 (2018), p. 2142–2155 | |
| dc.relation.references | Liu, Chen X. ; Chen, Y.: A hybrid convolutional neural network for multivariate time series anomaly detection. En: Information Sciences 565 (2021), p. 15–27 | |
| dc.relation.references | Liu, Ma T. Zhang X. Wang L. ; Li, X.: Deep Learning for Time Series Classification: A Review. En: Neurocomputing 338 (2019), p. 77–95 | |
| dc.relation.references | Lu, Han J. ; Wang, W.: A survey on multi-task learning. En: arXiv preprint ar- Xiv:1906.05022 (2019) | |
| dc.relation.references | Malhotra, Ramakrishnan A. A. ; G., L. Agarwal P. ; Shroff, G.: LSTM-based encoder-decoder for multi-sensor anomaly detection. En: Proceedings of the Internatio- nal Conference on Machine Learning (ICML), 2016, p. 2719–2728 | |
| dc.relation.references | Malhotra, Vig L. Shroff G. ; Agarwal, P.: Long Short Term Memory Networks for AnomalyDetectioninTimeSeries. En:Proceedings of the 23rd International Conference on Neural Information Processing (ICONIP), Springer, 2015, p. 1–10 | |
| dc.relation.references | Malhotra P, Shroff G Agarwal P.: LSTM-based encoder-decoder for multi-sensor anomaly detection. En: arXiv preprint (2016) | |
| dc.relation.references | Mirsky, Yisroel ; Doitshman, Tomer ; Elovici, Yuval ; Shabtai, Asaf: Kitsune: An ensemble of autoencoders for online network intrusion detection. arXiv 2018. En: arXiv preprint arXiv:1802.09089 (1802) | |
| dc.relation.references | Munir,Mohsin;Siddiqui,ShoaibA.;Dengel,Andreas;Ahmed,Sheraz: DeepAnT: A deep learning approach for unsupervised anomaly detection in time series. En: Ieee Access 7 (2018), p. 1991–2005 | |
| dc.relation.references | Nguyen, Nguyen H. T. Nguyen T. N. ; Phung, D.: A hybrid deep learning approach for stock price forecasting and anomaly detection. En: Expert Systems with Applications 116 (2019), p. 491–507 | |
| dc.relation.references | Pang, Li X. Li Z. ; Li, S.: TimeAutoAD: Autonomous Anomaly Detection With Self- Supervised Contrastive Loss for Multivariate Time Series. En: IEEE Transactions on Cybernetics 51 (2021), Nr. 3, p. 1381–1395 | |
| dc.relation.references | Pang, Shen C. Cao L. ; Hengel, A. V. D.: Deep learning for anomaly detection: A review. En: ACM Computing Surveys (CSUR) 54 (2021), Nr. 2, p. 1–38 | |
| dc.relation.references | Philip, F. M.: Video Anomaly Detection Using the Optimization-Enabled Deep Con- volutional Neural Network. En: The Computer Journal 65 (2022), Nr. 5, p. 1272–1292 | |
| dc.relation.references | Ren, He K. Girshick R. ; Sun, J.: Faster R-CNN: Towards real-time object detection with region proposal networks. En: Advances in neural information processing systems (2015), p. 91–99 | |
| dc.relation.references | Ruff, V ; ermeulen, Goossens A. ; Daelemans, W.: Semi-supervised anomaly detection in multivariate time series using LSTM encoder-decoder networks. En: Neural Networks 141 (2021), p. 42–53 | |
| dc.relation.references | Saleh, M. ; Bagheri, M.: Anomaly Detection in Financial Time Series Data Using Deep Autoencoder Neural Networks. En: Expert Systems with Applications 116 (2019), p. 139–153 | |
| dc.relation.references | Saleh, M. ; Eftekhari, M.: Anomaly detection in multivariate time series data using deep neural networks. En: Neurocomputing 332 (2019), p. 423–437 | |
| dc.relation.references | Sharma, A. ; Chen, T.: A deep learning approach for predicting stock price movement using numerical and textual information. En: Journal of Business Research 88 (2018), p. 1–16 | |
| dc.relation.references | Shyu, Chen S. C. Sarinnapakorn K. ; Chang, L.: A novel anomaly detection scheme based on principal component classifier. En: Proceedings of the 2003 IEEE International Conference on Data Mining, 2003, p. 167–174 | |
| dc.relation.references | Steane, Andrew: Quantum computing. En: Reports on Progress in Physics 61 (1998), Nr. 2, p. 117 | |
| dc.relation.references | Sun,LiuH.;Harada,T.: Onlinegrowingneuralgasforanomalydetectioninchanging surveillance scenes. En: Pattern Recognition 64 (2017), p. 187–201 | |
| dc.relation.references | Sun, Yang Y. Bo X. ; Zhao, J.: Hybrid deep learning-based spatiotemporal model for PM2. 5 concentration prediction. En: Applied Soft Computing 76 (2019), p. 180–191 | |
| dc.relation.references | Sun Y, Kamel M.: Classification of imbalanced data: A review. En: International Journal of Pattern Recognition and Artificial Intelligence 23 (2009), Nr. 4, p. 687–719 | |
| dc.relation.references | Särkkä, S.: Bayesian filtering and smoothing. Cambridge University Press, 2013 | |
| dc.relation.references | Wagner, Michels T. Schulz F. C. Nair A. Rudolph M. ; Kloft, M.: Timesead: Bench- marking deep multivariate time-series anomaly detection. En: Transactions on Machine Learning Research (2023) | |
| dc.relation.references | Wang, Wang Y. Zhang X. Li J. Li Y. ; Li, Y.: Anomaly detection for high-speed railway monitoring based on deep learning. En: IEEE Transactions on Instrumentation and Measurement 69 (2020), Nr. 4, p. 1464–1472 | |
| dc.relation.references | Xu, Hu W. Lu Z. ; Wen, J.: An LSTM-based deep learning method for anomaly detection in high-frequency financial time series. En: IEEE Access 9 (2021), p. 29655– 29666 | |
| dc.relation.references | Xu H, Mannor S.: Robust PCA via outlier pursuit. En: IEEE Transactions on Infor- mation Theory 56 (2010), Nr. 5, p. 2021–2037 | |
| dc.relation.references | Yang, Zhao L. Xie X. ; Liu, X.: Anomaly detection for smart building based on time series clustering and deep learning. En: Applied Soft Computing 95 (2020), p. 106569 | |
| dc.relation.references | Yuan,ZhengY.LuX.;Yang,C.: TimeSeriesAnomalyDetectionUsingConvolutional NeuralNetworksandLongShort-TermMemory. En:Proceedings of the 9th International Conference on Intelligent Systems and Knowledge Engineering (ISKE), IEEE, 2018, p. 47–54 | |
| dc.relation.references | Zhang, B. ; Song, X.: Time series anomaly detection with variational autoencoder. En: Proceedings of the 2018 International Joint Conference on Neural Networks, 2018, p. 1–8 | |
| dc.relation.references | Zhang, Kai ; Jiang, Yushan ; Seversky, Lee ; Xu, Chengtao ; Liu, Dahai ; Song, Houbing: Federated variational learning for anomaly detection in multivariate time series. En: 2021 IEEE International Performance, Computing, and Communications Conference (IPCCC) IEEE, 2021, p. 1–9 | |
| dc.relation.references | Zhang, Song D. Chen Y. Feng X. Lumezanu C. Cheng W. ; Chawla, N. V.: A deep neural network for unsupervised anomaly detection and diagnosis in multivariate time series data. En: Proceedings of the AAAI conference on artificial intelligence Vol. 33, 2019, p. 1409–1416 | |
| dc.relation.references | Zhang, Y. ; Hu, X.: A deep hybrid model for unsupervised anomaly detection in multivariate time series. En: Knowledge-Based Systems 181 (2019), p. 104832 | |
| dc.relation.references | Zhang G, Guo L Liu H.: A deep learning framework for financial time series using stacked autoencoders and long-short term memory. En: PLoS One 14 (2019), Nr. 5, p. e0216622 | |
| dc.relation.references | Zhao, Du S. Wang Y. Huang X. ; Zhou, M.: Multivariate Time Series Anomaly Detection Using a Deep Residual Autoencoder. En: Information Sciences 497 (2019), p. 132–142 | |
| dc.relation.references | Zhou, Li B. Li Y. ; Huang, W.: A Review of Convolutional Neural Networks for Time Series Analysis. En: Journal of Computer Science and Technology 35 (2020), Nr. 3, p. 545–563 | |
| dc.relation.references | Zhou, Z. H.: A brief introduction to weakly supervised learning. En: National Science Review 3 (2016), Nr. 1, p. 54–71 | |
| dc.rights.accessrights | info:eu-repo/semantics/openAccess | |
| dc.rights.license | Reconocimiento 4.0 Internacional | |
| dc.rights.uri | http://creativecommons.org/licenses/by/4.0/ | |
| dc.subject.ddc | 000 - Ciencias de la computación, información y obras generales::004 - Procesamiento de datos Ciencia de los computadores | |
| dc.subject.lemb | Análisis de series de tiempo | spa |
| dc.subject.lemb | Time-series analysis | eng |
| dc.subject.lemb | Aprendizaje automático (inteligencia artificial) | spa |
| dc.subject.lemb | Machine learning | eng |
| dc.subject.lemb | Decisiones estadísticas | spa |
| dc.subject.lemb | Statistical decision | eng |
| dc.subject.proposal | series temporales multivariantes | spa |
| dc.subject.proposal | detección de anomalías | spa |
| dc.subject.proposal | estimación de densidad | spa |
| dc.subject.proposal | aprendizaje profundo | spa |
| dc.subject.proposal | matrices de densidad | spa |
| dc.subject.proposal | inteligencia artificial | spa |
| dc.subject.proposal | datos secuenciales | spa |
| dc.title | Detección de anomalías en series temporales multivariantes (MTS) basada en métodos de estimación de la densidad. | spa |
| dc.title.translated | Anomaly detection in multivariate time series (MTS) based on density estimation methods. | eng |
| dc.type | Trabajo de grado - Maestría | |
| dc.type.coar | http://purl.org/coar/resource_type/c_bdcc | |
| dc.type.coarversion | http://purl.org/coar/version/c_ab4af688f83e57aa | |
| dc.type.content | Text | |
| dc.type.driver | info:eu-repo/semantics/masterThesis | |
| dc.type.redcol | http://purl.org/redcol/resource_type/TM | |
| dc.type.version | info:eu-repo/semantics/acceptedVersion | |
| dcterms.audience.professionaldevelopment | Estudiantes | |
| dcterms.audience.professionaldevelopment | Investigadores | |
| dcterms.audience.professionaldevelopment | Especializada | |
| oaire.accessrights | http://purl.org/coar/access_right/c_abf2 |