A method for clustering time-varying data by using neural networks, i.e. Kohonen self-organizing maps (SOMs), is suggested. Some dissimilarity measures for capturing the temporal structure of the data are introduced and used in Kohonen SOMs allowing clustering of temporal data. Another method for clustering time-varying data, called dynamic tandem analysis (DTA), based on the sequential utilization of dynamic factor analysis and cluster analysis, is also considered. The methods are applied to telecommunications market segmentation on real data. The obtained results are compared and discussed. (C) 2007 Elsevier B.V. All rights reserved.
Temporal self-organizing maps for telecommunications market segmentation / D'Urso, Pierpaolo; Livia De, Giovanni. - In: NEUROCOMPUTING. - ISSN 0925-2312. - 71:13-15(2008), pp. 2880-2892. [10.1016/j.neucom.2007.07.012]
Temporal self-organizing maps for telecommunications market segmentation
D'URSO, Pierpaolo;
2008
Abstract
A method for clustering time-varying data by using neural networks, i.e. Kohonen self-organizing maps (SOMs), is suggested. Some dissimilarity measures for capturing the temporal structure of the data are introduced and used in Kohonen SOMs allowing clustering of temporal data. Another method for clustering time-varying data, called dynamic tandem analysis (DTA), based on the sequential utilization of dynamic factor analysis and cluster analysis, is also considered. The methods are applied to telecommunications market segmentation on real data. The obtained results are compared and discussed. (C) 2007 Elsevier B.V. All rights reserved.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
