Conformal $k$-NN Anomaly Detector for Univariate Data Streams

Vladislav Ishimtsev, Alexander Bernstein, Evgeny Burnaev, Ivan Nazarov
Proceedings of the Sixth Workshop on Conformal and Probabilistic Prediction and Applications, PMLR 60:213-227, 2017.

Abstract

Anomalies in time-series data give essential and often actionable information in many applications. In this paper we consider a model-free anomaly detection method for univariate time-series which adapts to non-stationarity in the data stream and provides probabilistic abnormality scores based on the conformal prediction paradigm. Despite its simplicity the method performs on par with complex prediction-based models on the Numenta Anomaly Detection benchmark and the Yahoo! S5 dataset.

Cite this Paper

BibTeX
@InProceedings{pmlr-v60-ishimtsev17a, title = {Conformal $k$-{NN} Anomaly Detector for Univariate Data Streams}, author = {Ishimtsev, Vladislav and Bernstein, Alexander and Burnaev, Evgeny and Nazarov, Ivan}, booktitle = {Proceedings of the Sixth Workshop on Conformal and Probabilistic Prediction and Applications}, pages = {213--227}, year = {2017}, editor = {Gammerman, Alex and Vovk, Vladimir and Luo, Zhiyuan and Papadopoulos, Harris}, volume = {60}, series = {Proceedings of Machine Learning Research}, month = {13--16 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v60/ishimtsev17a/ishimtsev17a.pdf}, url = {https://proceedings.mlr.press/v60/ishimtsev17a.html}, abstract = {Anomalies in time-series data give essential and often actionable information in many applications. In this paper we consider a model-free anomaly detection method for univariate time-series which adapts to non-stationarity in the data stream and provides probabilistic abnormality scores based on the conformal prediction paradigm. Despite its simplicity the method performs on par with complex prediction-based models on the Numenta Anomaly Detection benchmark and the Yahoo! S5 dataset.} }
Endnote
%0 Conference Paper %T Conformal $k$-NN Anomaly Detector for Univariate Data Streams %A Vladislav Ishimtsev %A Alexander Bernstein %A Evgeny Burnaev %A Ivan Nazarov %B Proceedings of the Sixth Workshop on Conformal and Probabilistic Prediction and Applications %C Proceedings of Machine Learning Research %D 2017 %E Alex Gammerman %E Vladimir Vovk %E Zhiyuan Luo %E Harris Papadopoulos %F pmlr-v60-ishimtsev17a %I PMLR %P 213--227 %U https://proceedings.mlr.press/v60/ishimtsev17a.html %V 60 %X Anomalies in time-series data give essential and often actionable information in many applications. In this paper we consider a model-free anomaly detection method for univariate time-series which adapts to non-stationarity in the data stream and provides probabilistic abnormality scores based on the conformal prediction paradigm. Despite its simplicity the method performs on par with complex prediction-based models on the Numenta Anomaly Detection benchmark and the Yahoo! S5 dataset.
APA
Ishimtsev, V., Bernstein, A., Burnaev, E. & Nazarov, I.. (2017). Conformal $k$-NN Anomaly Detector for Univariate Data Streams. Proceedings of the Sixth Workshop on Conformal and Probabilistic Prediction and Applications, in Proceedings of Machine Learning Research 60:213-227 Available from https://proceedings.mlr.press/v60/ishimtsev17a.html.

Related Material