Domain agnostic online semantic segmentation for multi-dimensional time series

Abstract

Unsupervised semantic segmentation in the time series domain is a much studied problem due to its potential to detect unexpected regularities and regimes in poorly understood data. However, the current techniques have several shortcomings, which have limited the adoption of time series semantic segmentation beyond academic settings for four primary reasons. First, most methods require setting/learning many parameters and thus may have problems generalizing to novel situations. Second, most methods implicitly assume that all the data is segmentable and have difficulty when that assumption is unwarranted. Thirdly, many algorithms are only defined for the single dimensional case, despite the ubiquity of multi-dimensional data. Finally, most research efforts have been confined to the batch case, but online segmentation is clearly more useful and actionable. To address these issues, we present a multi-dimensional algorithm, which is domain agnostic, has only one, easily-determined parameter, and can handle data streaming at a high rate. In this context, we test the algorithm on the largest and most diverse collection of time series datasets ever considered for this task and demonstrate the algorithm’s superiority over current solutions.

Visibility graph based temporal community detection with applications in biological time series

“…We have implemented a graph-based method to explore the community detection for graphs representing time series, building on previously established visibility graph approaches. There are also many other algorithms utilizing analytic methods for similar tasks, for example segmentation algorithms 53 , 54 . A notable recent implementation is the FLOSS (Fast Low-cost Online Semantic Segmentation) algorithm, which also is visually similar to our approach, using ‘Arcs’ as connectivity variables, in contrast with our method using visibility connections 54 .…”

“…There are also many other algorithms utilizing analytic methods for similar tasks, for example segmentation algorithms 53 , 54 . A notable recent implementation is the FLOSS (Fast Low-cost Online Semantic Segmentation) algorithm, which also is visually similar to our approach, using ‘Arcs’ as connectivity variables, in contrast with our method using visibility connections 54 . Since we have taken a graph-based approach, we use graph terminology: for example we refer to ‘temporal community detection’, instead of ‘regime discovery’ or ‘semantic segmentation’ which may be considered as analogous constructions, though with rather different underlying methodologies.…”

Bayesian Online Prediction of Change Points

“…The segmentation and classification of time series has many applications in different fields such as predicting failures in an oil process plant, reconstructing trajectories in air traffic control, the identification of interaction scenarios in robotic environments, real-time brain computer interfaces, probabilistic forecasting of volcano eruptions or forecasting time series with multiple seasonal patterns (Molina et al, 2009;Gharghabi et al, 2018;Xiao et al, 2018;Cassisi et al, 2016;Gould et al, 2008). For a recent review of change point detection methods we refer to Aminikhanghahi & Cook (2017); Truong et al (2018).…”

Hybrid Algorithm for Anomaly Removal in Time Series Data Mining

“…Data mining professionals tend to use fast, unsupervised methods in the early stages of the data mining process. Thus, data mining tasks (e.g., motif discovery, discord discovery, clustering, and segmentation) should be handled for time series data [4][5][6][7][8]. However, time series have anomalies due to similarities [9,10].…”