Exploiting self-similarity for change detection

Giacomo Boracchi, Manuel Roveri

Abstract

Time-series data are often characterized by a large degree of self-similarity, which arises in application domains featuring periodicity or seasonality. While self-similarity has shown to be an effective prior for modeling real data in the signal and image-processing literature, it has received much less attention in time-series literature, where only few works leveraging the self-similarity for anomaly detection have been presented. Here we introduce a novel change-detection test to detect structural changes in time series by analyzing their self-similarity. The core of the proposed solution is the definition of a change indicator to quantitatively assesses the self-similarity of the time-series data over time. In particular, the change indicator is obtained by comparing each patch to be analyzed with its most similar counterpart in a change-free training set. Experimental results on the flow measurements in the water distribution network of the Barcelona city show the effectiveness of the proposed solution.
BibTeX
@inproceedings{boracchi2014exploiting,
  author = {Giacomo Boracchi and Manuel Roveri},
  title = {{Exploiting self-similarity for change detection}},
  booktitle = {IJCNN},
  year = {2014},
  doi = {10.1109/ijcnn.2014.6889860},
}

Loading related content...

Figure