Level change detection in time series using higher order statistics

C. S. Hilas; I. T. Rekanos; S. K. Goudos; P. A. Mastorocostas; J. N. Sahalos

doi:10.1109/ICDSP.2009.5201260

Level change detection in time series using higher order statistics

C. S. Hilas, I. T. Rekanos, S. K. Goudos, P. A. Mastorocostas, J. N. Sahalos

Department of Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Changes in the level of a time series are usually attributed to an intervention that interrupts its evolution. The resulting time series are referred to as interrupted time series and they are studied in order to measure, e.g. the impact of new laws or medical treatments. In the present paper a heuristic method for level change detection in non-stationary time series is presented. The method uses higher order statistics, namely the skewness and the kurtosis, and can identify both the existence of a change in the level of the time series as well as the time point it has happened. The technique is tested with both simulated and real world data and is straightforward applicable to the detection of outliers in time series.

Original language	English
Title of host publication	DSP 2009
Subtitle of host publication	16th International Conference on Digital Signal Processing, Proceedings
DOIs	https://doi.org/10.1109/ICDSP.2009.5201260
Publication status	Published - 2009
Event	DSP 2009:16th International Conference on Digital Signal Processing - Santorini, Greece Duration: 5 Jul 2009 → 7 Jul 2009

Other

Other	DSP 2009:16th International Conference on Digital Signal Processing
Country/Territory	Greece
City	Santorini
Period	5/07/09 → 7/07/09

Keywords

Change point detection
Heuristic methods
Higher order statistics
Interrupted time series

Access to Document

10.1109/ICDSP.2009.5201260

Cite this

@inproceedings{c3b981ed7ea64f2cac81d6ffcb16df28,

title = "Level change detection in time series using higher order statistics",

abstract = "Changes in the level of a time series are usually attributed to an intervention that interrupts its evolution. The resulting time series are referred to as interrupted time series and they are studied in order to measure, e.g. the impact of new laws or medical treatments. In the present paper a heuristic method for level change detection in non-stationary time series is presented. The method uses higher order statistics, namely the skewness and the kurtosis, and can identify both the existence of a change in the level of the time series as well as the time point it has happened. The technique is tested with both simulated and real world data and is straightforward applicable to the detection of outliers in time series.",

keywords = "Change point detection, Heuristic methods, Higher order statistics, Interrupted time series",

author = "Hilas, {C. S.} and Rekanos, {I. T.} and Goudos, {S. K.} and Mastorocostas, {P. A.} and Sahalos, {J. N.}",

year = "2009",

doi = "10.1109/ICDSP.2009.5201260",

language = "English",

isbn = "9781424432981",

booktitle = "DSP 2009",

note = "DSP 2009:16th International Conference on Digital Signal Processing ; Conference date: 05-07-2009 Through 07-07-2009",

}

Hilas, CS, Rekanos, IT, Goudos, SK, Mastorocostas, PA & Sahalos, JN 2009, Level change detection in time series using higher order statistics. in DSP 2009: 16th International Conference on Digital Signal Processing, Proceedings., 5201260, DSP 2009:16th International Conference on Digital Signal Processing, Santorini, Greece, 5/07/09. https://doi.org/10.1109/ICDSP.2009.5201260

TY - GEN

T1 - Level change detection in time series using higher order statistics

AU - Hilas, C. S.

AU - Rekanos, I. T.

AU - Goudos, S. K.

AU - Mastorocostas, P. A.

AU - Sahalos, J. N.

PY - 2009

Y1 - 2009

N2 - Changes in the level of a time series are usually attributed to an intervention that interrupts its evolution. The resulting time series are referred to as interrupted time series and they are studied in order to measure, e.g. the impact of new laws or medical treatments. In the present paper a heuristic method for level change detection in non-stationary time series is presented. The method uses higher order statistics, namely the skewness and the kurtosis, and can identify both the existence of a change in the level of the time series as well as the time point it has happened. The technique is tested with both simulated and real world data and is straightforward applicable to the detection of outliers in time series.

AB - Changes in the level of a time series are usually attributed to an intervention that interrupts its evolution. The resulting time series are referred to as interrupted time series and they are studied in order to measure, e.g. the impact of new laws or medical treatments. In the present paper a heuristic method for level change detection in non-stationary time series is presented. The method uses higher order statistics, namely the skewness and the kurtosis, and can identify both the existence of a change in the level of the time series as well as the time point it has happened. The technique is tested with both simulated and real world data and is straightforward applicable to the detection of outliers in time series.

KW - Change point detection

KW - Heuristic methods

KW - Higher order statistics

KW - Interrupted time series

UR - http://www.scopus.com/inward/record.url?scp=70449638104&partnerID=8YFLogxK

U2 - 10.1109/ICDSP.2009.5201260

DO - 10.1109/ICDSP.2009.5201260

M3 - Conference contribution

AN - SCOPUS:70449638104

SN - 9781424432981

BT - DSP 2009

T2 - DSP 2009:16th International Conference on Digital Signal Processing

Y2 - 5 July 2009 through 7 July 2009

ER -

Level change detection in time series using higher order statistics

Abstract

Other

Keywords

Access to Document

Other files and links

Fingerprint

Cite this