Discord Aware Matrix Profile

Scaling Time Series Anomaly Detection to Trillions of Datapoints and Ultra-fast Arriving Data Streams

DAMP appears in: Matrix Profile XXIV:Scaling Time Series Anomaly Detection to Trillions of Datapoints and Ultra-fast Arriving Data Streams. Yue Lu, Renjie Wu, Abdullah Mueen, Maria A. Zuluaga and Eamonn Keogh. ACM SIGKDD 2022 [PDF] [Slides] [Poster] [Video Presentation]

DAMP_vid.mp4

Real-time Time Series Anomaly Search

This video shows time series anomaly search in real time, on a cheap commodity computer: Intel® Core™ i5-8500 CPU @ 3.00 GHz

The dataset is:

  • Eight million datapoints long

  • 17.36 hours of wall clock time

  • 85,056 heartbeats, which have been annotated by a combination of algorithms and cardiologists.


The search took 22.3 seconds, which is 2,802 times faster than real-time.