Influence function of projection-pursuit principal components for functional data

In the finite-dimensional setting, Li and Chen (1985) proposed a method for principal components analysis using projection-pursuit techniques. This procedure was generalized to the functional setting by Bali et al. (2011), where also different penalized estimators were defined to provide smooth func...

Descripción completa

Detalles Bibliográficos
Autores: Bali, Juan Lucas, Boente Boente, Graciela Lina
Tipo de recurso: artículo
Estado:Versión publicada
Fecha de publicación:2015
País:Argentina
Institución:Consejo Nacional de Investigaciones Científicas y Técnicas
Repositorio:CONICET Digital (CONICET)
Idioma:inglés
OAI Identifier:oai:ri.conicet.gov.ar:11336/18939
Acceso en línea:http://hdl.handle.net/11336/18939
Access Level:acceso abierto
Palabra clave:Elliptical Distribution
Fisher-Consistency
Functional Principal Component
Influence Function
Robust Estimation
Smoothing
https://purl.org/becyt/ford/1.1
https://purl.org/becyt/ford/1
Descripción
Sumario:In the finite-dimensional setting, Li and Chen (1985) proposed a method for principal components analysis using projection-pursuit techniques. This procedure was generalized to the functional setting by Bali et al. (2011), where also different penalized estimators were defined to provide smooth functional robust principal component estimators. This paper completes their study by deriving the influence function of the functional related to the principal direction estimators and their size. As is well known, the influence function is a measure of robustness which can also be used for diagnostic purposes. In this sense, the obtained results can be helpful for detecting influential observations for the principal directions.