Functional principal component analysis of quantile curves

Risultato della ricerca: Other

Abstract

Literature on functional data analysis is mainly focused on estimation ofindividuals curves and characterization of average dynamics. The idea underlyingthis proposal is to focus attention on other particular features of the distribution ofthe observed data, moving from mean functions towards functional quantiles. Themotivating examples are functional data sets that are collections of high frequencydata recorded along time. As quantiles provide information on various aspects of atime series, we propose a modelling framework for the joint estimation of functionalquantiles, varying along time, and functional principal components, summarizingsome common dynamics shared by the functional quantiles.
Lingua originaleEnglish
Pagine887-892
Numero di pagine6
Stato di pubblicazionePublished - 2017

    Fingerprint

Cita questo