A two-level algorithm of time series change detection based on a unique changes similarity method

In the paper, a novel two level algorithm of time series change detection is presented. In the first level, to identify non-stationary sequences in processed signals preliminary detection of events is performed with short-term prediction comparison. In the second stage, to confirm changes detected in first level a unique changes similarity method is employed. Detection of changes in non-stationary time series is discussed, implemented algorithms are described and results produced on exemplary four financial time series are showed.