Mining for weak periodic signals in time series databases

Article type: Research Article

Authors: Berberidis, Christos^{; *} | Vlahavas, Ioannis

Affiliations: Department of Informatics, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece

Correspondence: [*] Corresponding author: Christos Berberidis, Department of Informatics, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece. Tel.: +30 231 099 8418; Fax: +30 231 099 8362, E-mail: [email protected]

Abstract: Periodicity is a particularly interesting feature, which is often inherent in real world time series data sets. In this article we propose a data mining technique for detecting multiple partial and approximate periodicities. Our approach is exploratory and follows a filter/refine paradigm. In the filter phase we introduce an autocorrelation-based algorithm that produces a set of candidate partial periodicities. The algorithm is extended to capture approximate periodicities. In the refine phase we effectively prune invalid periodicities. We conducted a series of experiments with various real-world data sets to test the performance and verify the quality of the results.

Keywords: time series mining, periodicity detection

DOI: 10.3233/IDA-2005-9103

Journal: Intelligent Data Analysis, vol. 9, no. 1, pp. 29-42, 2005