Exploratory analysis of functional data via clustering and optimal segmentation

• Published in: Neurocomputing (Amsterdam) Vol. 73; no. 7; pp. 1125 - 1141
• Main Authors: Hébrail, Georges, Hugueney, Bernard, Lechevallier, Yves, Rossi, Fabrice
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01-03-2010; Elsevier
• Subjects: Clustering; Computer Science; Dynamic programming; Exploratory analysis; Functional data; Machine Learning; Multiple time series; Other Statistics; Segmentation; Statistics; Exploratory analysis; Dynamic programming; Functional data; Segmentation; Clustering; Multiple time series; Functional Data
• ISSN: 0925-2312; 1872-8286
• DOI: 10.1016/j.neucom.2009.11.022

We propose in this paper an exploratory analysis algorithm for functional data. The method partitions a set of functions into
K clusters and represents each cluster by a simple prototype (e.g., piecewise constant). The total number of segments in the prototypes,
P, is chosen by the user and optimally distributed among the clusters via two dynamic programming algorithms. The practical relevance of the method is shown on two real world datasets.