RT Journal Article
T1 Exact risk improvement of bandwidth selectors for kernel density estimation with directional data
A1 García Portugués, Eduardo
K1 Bandwidth selection
K1 Directional data
K1 Mixtures
K1 Kernel density estimator
K1 Von Mises
AB New bandwidth selectors for kernel density estimation withdirectional data are presented in this work. These selectors are based onasymptotic and exact error expressions for the kernel density estimatorcombined with mixtures of von Mises distributions. The performance ofthe proposed selectors is investigated in a simulation study and comparedwith other existing rules for a large variety of directional scenarios, samplesizes and dimensions. The selector based on the exact error expression turnsout to have the best behaviour of the studied selectors for almost all thesituations. This selector is illustrated with real data for the circular andspherical cases
PB The Institute of Mathematical Statistics
PB The Bernoulli Society
YR 2013
FD 2013
LK http://hdl.handle.net/10347/18565
UL http://hdl.handle.net/10347/18565
LA eng
NO García–Portugués, Eduardo. Exact risk improvement of bandwidth selectors for kernel density estimation with directional data. Electron. J. Statist., 7, 1655-1685. doi:10.1214/13-EJS821
NO The work of the author has been supported by FPU grant AP2010–0957 from the SpanishMinistry of Education. Support of Project MTM2008–03010, from the Spanish Ministry ofScience and Innovation, Project 10MDS207015PR from Dirección Xeral de I+D, Xunta deGalicia and IAP network StUDyS, from Belgian Science Policy, are acknowledged
DS Minerva
RD 30 abr 2026