Advances in functional regression and classification models

Functional data analysis (FDA) has become a very active field of research in the last few years because it appears naturally in most scientific fields: energy (electricity price curves), environment (curves of pollutant levels), chemometrics (spectrometric data), etc. This thesis is a compendium of...

ver descrição completa

Detalhes bibliográficos
Autor: Oviedo de la Fuente, Manuel
Formato: tesis doctoral
Fecha de publicación:2019
País:España
Recursos:Universidad de Santiago de Compostela (USC)
Repositorio:Minerva. Repositorio Institucional de la Universidad de Santiago de Compostela
Idioma:inglés
OAI Identifier:oai:minerva.usc.gal:10347/18236
Acesso em linha:http://hdl.handle.net/10347/18236
Access Level:acceso abierto
Palavra-chave:Materias::Investigación::12 Matemáticas::1209 Estadística::120914 Técnicas de predicción estadística
Materias::Investigación::12 Matemáticas::1209 Estadística::120903 Análisis de datos
Descrição
Resumo:Functional data analysis (FDA) has become a very active field of research in the last few years because it appears naturally in most scientific fields: energy (electricity price curves), environment (curves of pollutant levels), chemometrics (spectrometric data), etc. This thesis is a compendium of the following publications: 1) "Statistical computing in functional data analysis: the R package fda.usc" published in the J STAT SOFTW, the core advances of this paper was to propose a common framework for FDA in R. 2) "Predicting seasonal influenza transmission using functional regression models with temporal dependence" published in PLoS ONE proposes an extension of GLS model to functional case. 3) "The DD$^G$--classifier in the functional setting" published in TEST extends the DD-classifier using information derived of the functional depth. 4) "Determining optimum wavelengths for leaf water content estimation from reflectance: A distance correlation approach" published in CHEMOMETR INTELL LAB SYST studies the utility of distance correlation as a method to select impact points in functional regression. 5) "Variable selection in Functional Additive Regression Models", in Comput Stat proposes a variable selection algorithm in the case of mixed predictors (scalar, functional, etc.).