Online Amnesic Approximation of Streaming Time Series

  • Authors:
  • Themistoklis Palpanas;Michail Vlachos;Eamonn Keogh;Dimitrios Gunopulos;Wagner Truppel

  • Affiliations:
  • -;-;-;-;-

  • Venue:
  • ICDE '04 Proceedings of the 20th International Conference on Data Engineering
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

The past decade has seen a wealth of research on time series representations, because the manipulation, storage, andindexing of large volumes of raw time series data is impractical. The vast majority of research has concentrated on representations that are calculated in batch mode and representeach value with approximately equal fidelity. However, the increasing deployment of mobile devices and real time sensorshas brought home the need for representations that can beincrementally updated, and can approximate the data with fidelity proportional to its age. The latter property allows us toanswer queries about the recent past with greater precision,since in many domains recent information is more useful thanolder information. We call such representations amnesic.While there has been previous work on amnesic representations, the class of amnesic functions possible was dictatedby the representation itself. In this work, we introduce anovel representation of time series that can represent arbitrary, user-specified amnesic functions. For example, a meteorologist may decide that data that is twice as old can toleratetwice as much error, and thus, specify a linear amnesic function. In contrast, an econometrist might opt for an exponentialamnesic function. We propose online algorithms for our representation, and discuss their properties. Finally, we performan extensive empirical evaluation on 40 datasets, and showthat our approach can efficiently maintain a high quality amnesicapproximation.