Construction of Data Models by Means of Hierarchical Archimedean Copulas

Thumbnail Image

Files

GOR0013_FEI_P1807_1801V001_2015.pdf (2.56 MB)
GOR0013_FEI_P1807_1801V001_2015_autoreferat.pdf (587.91 KB)
GOR0013_FEI_P1807_1801V001_2015_posudek_oponent_Antoch_Jaromir.pdf (149.53 KB)
GOR0013_FEI_P1807_1801V001_2015_posudek_oponent_Hofert_Marius.pdf (47.83 KB)
GOR0013_FEI_P1807_1801V001_2015_posudek_oponent_Ramik_Jaroslav.pdf (400.16 KB)

Downloads

17

Date issued

2015

Authors

Górecki, Jan

Journal Title

Journal ISSN

Volume Title

Publisher

Vysoká škola báňská - Technická univerzita Ostrava

Location

ÚK/Sklad diplomových prací

Signature

201500932

Abstract

This dissertation work explores various approaches to hierarchical Archimedean copula estimation. In particular, efficient estimation algorithms are proposed and experimentally tested on simulated data. As an application, a new set of Bayesian classifiers based on hierarchical Archimedean copulas is constructed and experimentally compared with other popular classifiers in terms of accuracy.

Description

Import 04/11/2015
Import 02/11/2016

Subject(s)

copula, hierarchical Archimedean copula, copula estimation, structure determination, Archimedean family of copulas, Kendall's tau, goodness-of-fit, Bayesian classification

Citation

Item identifier

http://hdl.handle.net/10084/110919

Collections

Vysokoškolské kvalifikační práce Fakulty elektrotechniky a informatiky / Theses and dissertations of Faculty of Electrical Engineering and Computer Science (FEI)