Corpus ID: 5904879

Conformal k-NN Anomaly Detector for Univariate Data Streams

@inproceedings{Ishimtsev2017ConformalKA,
  title={Conformal k-NN Anomaly Detector for Univariate Data Streams},
  author={V. Ishimtsev and A. Bernstein and Evgeny Burnaev and I. Nazarov},
  booktitle={COPA},
  year={2017}
}
  • V. Ishimtsev, A. Bernstein, +1 author I. Nazarov
  • Published in COPA 2017
  • Computer Science, Mathematics
  • 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. 
    Learning Ensembles of Anomaly Detectors on Synthetic Data
    5
    Exchangeability martingales for selecting features in anomaly detection
    Usage of multiple RTL features for Earthquake prediction
    1
    Deep Ensembles for Imbalanced Classification
    8
    Rare Failure Prediction via Event Matching for Aerospace Applications
    1
    Sequence embeddings help to identify fraudulent cases in healthcare insurance
    3

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 46 REFERENCES
    Conformalized density- and distance-based anomaly detection in time-series data
    11
    Inductive conformal anomaly detection for sequential detection of anomalous sub-trajectories
    18
    Time Series Clustering for Anomaly Detection Using Competitive Neural Networks
    24
    Evaluating Real-Time Anomaly Detection Algorithms -- The Numenta Anomaly Benchmark
    158
    Outlier Detection Using Replicator Neural Networks
    485
    Time-series novelty detection using one-class support vector machines
    218
    Model selection for anomaly detection
    22
    Outlier Detection in Multivariate Time Series by Projection Pursuit
    94
    On-line novelty detection using the Kalman filter and extreme value theory
    43