@INPROCEEDINGS{boracchiIJCNN14_SelfSimilarity, 
author={Boracchi, G. and Roveri, M.}, 
booktitle={Neural Networks (IJCNN), 2014 International Joint Conference on}, 
title={Exploiting self-similarity for change detection}, 
year={2014}, 
month={July}, 
pages={3339-3346}, 
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.}, 
keywords={data analysis;time series;Barcelona city;change indicator;change-detection test;change-free training set;flow measurements;time series structural change detection;time-series data self-similarity;water distribution network;Correlation;Monitoring;Predictive models;Random variables;Time measurement;Time series analysis;Training}, 
doi={10.1109/IJCNN.2014.6889860},}