RT info:eu-repo/semantics/article
T1 Time series clustering using the the total variation distance with applications in Oceanography
A1 Álvarez Esteban, Pedro César
A1 Euán, C.
A1 Ortega, J.
K1 Oceanografía - Estadística
AB A time series clustering algorithm based on the use of the total variation distancebetween normalized spectra as a measure of dissimilarity is proposed in this work.The oscillatory behavior of the series is thus considered the central characteristic forclassi cation purposes. The proposed algorithm is compared to several other methodswhich are also based on features extracted from the original series and the results showthat its performance is comparable to the best methods available and in some tests itoutperforms the rest. As an application the algorithm is used to determine stationaryperiods for random sea waves, both in simulations and on a real data set, a problemin which changes between stationary sea states are usually slow.
PB Universidad de Valladolid. Facultad de Medicina
YR 2015
FD 2015
LK http://uvadoc.uva.es/handle/10324/8296
UL http://uvadoc.uva.es/handle/10324/8296
LA eng
NO Arxiv, 21 jav. 2015 p.1-23
NO Producción Científica
DS UVaDOC
RD 12-sep-2024