Constrained distance based clustering for time-series: a comparative and experimental study

• Published in: Data mining and knowledge discovery Vol. 32; no. 6; pp. 1663 - 1707
• Main Authors: Lampert, Thomas, Dao, Thi-Bich-Hanh, Lafabregue, Baptiste, Serrette, Nicolas, Forestier, Germain, Crémilleux, Bruno, Vrain, Christel, Gançarski, Pierre
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.11.2018; Springer Nature B.V; Springer
• Subjects: Academic Surveys and Tutorials; Algorithms; Artificial Intelligence; Chemistry and Earth Sciences; Clustering; Comparative studies; Computer Science; Constraints; Data mining; Data Mining and Knowledge Discovery; Distance measurement; Euclidean geometry; Information Storage and Retrieval; Physics; Statistics for Engineering; Time series; Partition clustering; Time-series; Semi-supervised clustering; Constrained clustering; Dynamic time warping
• ISSN: 1384-5810; 1573-756X; 1573-756X
• DOI: 10.1007/s10618-018-0573-y

Constrained clustering is becoming an increasingly popular approach in data mining. It offers a balance between the complexity of producing a formal definition of thematic classes—required by supervised methods—and unsupervised approaches, which ignore expert knowledge and intuition. Nevertheless, the application of constrained clustering to time-series analysis is relatively unknown. This is partly due to the unsuitability of the Euclidean distance metric, which is typically used in data mining, to time-series data. This article addresses this divide by presenting an exhaustive review of constrained clustering algorithms and by modifying publicly available implementations to use a more appropriate distance measure—dynamic time warping. It presents a comparative study, in which their performance is evaluated when applied to time-series. It is found that k -means based algorithms become computationally expensive and unstable under these modifications. Spectral approaches are easily applied and offer state-of-the-art performance, whereas declarative approaches are also easily applied and guarantee constraint satisfaction. An analysis of the results raises several influencing factors to an algorithm’s performance when constraints are introduced.