Empirical likelihood in Euclidean and some non-Euclidean spaces

Yan, Xi (2020) Empirical likelihood in Euclidean and some non-Euclidean spaces. PhD thesis, University of Nottingham.

[img] PDF (Thesis - as examined) - Repository staff only - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Download (2MB)

Abstract

Empirical likelihood is a non-parametric approach to statistical inference which bears some resemblance to parametric approaches and yet avoids making parametric assumptions. The aim of this thesis is to develop theoretical and computational aspects of empirical likelihood when applied to data from two different types of non-Euclidean space. The two spaces we focus on are the sphere, which is an important example of a manifold with non-Euclidean structure and is the appropriate sample space in directional statistics; and the 3-Spider, which is an example of a stratified manifold. The focus is mainly on developing empirical likelihood for a Fréchet mean in these two settings. Key achievements are: to prove versions of Wilks’ theorem which open the way to the application of methods of inference including the construction of confidence regions and hypothesis testing; and the development of algorithms for performing the computations.

Item Type: Thesis (University of Nottingham only) (PhD)
Supervisors: Wood, Andrew
Le, Huiling
Keywords: Empirical Likelihood; Wilks' Theorem for Empirical Likelihood; Local Alternatives; Concavity; Directional Data; Sphere; Fréchet Mean; 3-Spider; Phylogenetic Tree Data; Confidence Region
Subjects: Q Science > QA Mathematics > QA440 Geometry
Faculties/Schools: UK Campuses > Faculty of Science > School of Mathematical Sciences
Item ID: 59820
Depositing User: Yan, Xi
Date Deposited: 15 Jul 2020 04:40
Last Modified: 15 Jul 2020 04:40
URI: http://eprints.nottingham.ac.uk/id/eprint/59820

Actions (Archive Staff Only)

Edit View Edit View