A generalized matrix profile framework with support for contextual series analysis
作者： Dieter De Paepe , Sander Vanden Hautte , Bram Steenwinckel , Filip De Turck , Femke Ongenae
DOI: 10.1016/J.ENGAPPAI.2020.103487
关键词: Data mining 、 Anomaly detection 、 Distance matrix 、 Computer science 、 Time series 、 Distance measures 、 Matrix (mathematics)
摘要: Abstract The Matrix Profile is a state-of-the-art time series analysis technique that can be used for motif discovery, anomaly detection, segmentation and others, in various domains such as healthcare, robotics, audio. Where recent techniques use the preprocessing or modeling step, we believe there unexplored potential generalizing approach. We derived framework focuses on implicit distance matrix calculation. present this Series Distance (SDM). In framework, measures (SDM-generators) processors (SDM-consumers) freely combined, allowing more flexibility easier experimentation. SDM, but one specific configuration. also introduce Contextual (CMP) new SDM-consumer capable of discovering repeating patterns. CMP provides intuitive visualizations data find anomalies are not discords. demonstrate using two real world cases. first wide variety fits within SDM complement Profile.
sci-hub.se 参考文章(24)
Diego Furtado Silva, Gustavo EAPA Batista, None, Elastic Time Series Motifs and Discords international conference on machine learning and applications. pp. 237- 242 ,(2018) , 10.1109/ICMLA.2018.00042
Dieter De Paepe, Olivier Janssens, Sofie Van Hoecke, Eliminating noise in the matrix profile international conference on pattern recognition applications and methods. ,vol. 2, pp. 83- 93 ,(2019) , 10.5220/0007314100830093
Dieter De Paepe, Diego Nieves Avendano, Sofie Van Hoecke, Implications of Z-Normalization in the Matrix Profile international conference on pattern recognition applications and methods. ,vol. 11996, pp. 95- 118 ,(2019) , 10.1007/978-3-030-40014-9_5
Nitin Kumar, Venkata Nishanth Lolla, Eamonn Keogh, Stefano Lonardi, Chotirat (Ann) Ratanamahatana, Li Wei, Time-series Bitmaps: a Practical Visualization Tool for Working with Large Time Series Databases. siam international conference on data mining. pp. 531- 535 ,(2005)
Haemwaan Sivaraks, Chotirat Ann Ratanamahatana, Robust and accurate anomaly detection in ECG artifacts using time series motif discovery. Computational and Mathematical Methods in Medicine. ,vol. 2015, pp. 453214- 453214 ,(2015) , 10.1155/2015/453214
Xiaopeng Xi, Michail Vlachos, Eamonn Keogh, Li Wei, Sang-Hee Lee, LB_Keogh supports exact indexing of shapes under rotation invariance with arbitrary representations and distance measures very large data bases. pp. 882- 893 ,(2006) , 10.5555/1182635.1164203
Varun Chandola, Arindam Banerjee, Vipin Kumar, Anomaly detection: A survey ACM Computing Surveys. ,vol. 41, pp. 15- ,(2009) , 10.1145/1541880.1541882
Alexander Lavin, Subutai Ahmad, Evaluating Real-Time Anomaly Detection Algorithms -- The Numenta Anomaly Benchmark international conference on machine learning and applications. pp. 38- 44 ,(2015) , 10.1109/ICMLA.2015.141
D. Vries, B. van den Akker, E. Vonk, W. de Jong, J. van Summeren, Application of machine learning techniques to predict anomalies in water supply networks Water Science & Technology: Water Supply. ,vol. 16, pp. 1528- 1535 ,(2016) , 10.2166/WS.2016.062
Chin-Chia Michael Yeh, Yan Zhu, Liudmila Ulanova, Nurjahan Begum, Yifei Ding, Hoang Anh Dau, Diego Furtado Silva, Abdullah Mueen, Eamonn Keogh, Matrix Profile I: All Pairs Similarity Joins for Time Series: A Unifying View That Includes Motifs, Discords and Shapelets 2016 IEEE 16th International Conference on Data Mining (ICDM). pp. 1317- 1322 ,(2016) , 10.1109/ICDM.2016.0179