Tests de independencia entre obxectos estatísticos
Loading...
Identifiers
Publication date
Authors
Advisors
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
[GL] Examínase cunha linguaxe marcada pola análise funcional e a teoría da medida o
problema matemático da independencia estatística dende unha perspectiva contemporánea.
Introdúcese o marco conceptual preciso de cara a acadar tal obxectivo para posteriormente
desenvolver de xeito rigoroso os dous formalismos que historicamente guiaron a evolución
do estudo da independencia, así como o sorprendente paralelismo que entre eles gardan.
Transpasando a fronteira entre o mundo poboacional e o empírico, preséntase de xeito
conciso un corpus de resultados relativos a U-estatísticos, abstracción fundamental da
teoría da estimación. A finalidade deste capítulo é proporcionar a maquinaria precisa para
dar paso á formulación empírica do contraste de independencia de dúas variables aleatorias
dada unha mostra. Finalmente, condénsanse as contribucións orixinais nos dous derradeiros
capítulos proporcionando unha ponte entre o exposto até este punto e a teoría de machine
learning máis actual, para así dar paso á aplicación do presente traballo no eido da física
de materiais e a novedosa representación da súa estrutura en forma de grafos.
[EN] The mathematical problem of statistical independence is examined from a contemporary perspective with a language marked by functional analysis and measure theory. We introduce the necessary conceptual framework in order to achieve this objective and then rigorously develop the two formalisms that historically guided the evolution of the study of independence, as well as the surprising parallelism between them. Crossing the border between the population and empirical worlds, we present in a concise way a corpus of results related to U-statistics, a fundamental abstraction of estimation theory. The purpose of this chapter is to provide the machinery needed to give way to the empirical formulation of testing given a sample whether two random variables are independent or not. Finally, the original contributions are condensed in the two following chapters, providing a bridge between what has been exposed up to this point and the most current machine learning theory, in order to show the applicability of this work in the field of materials physics and the novel representation of their structure in the form of graphs
[EN] The mathematical problem of statistical independence is examined from a contemporary perspective with a language marked by functional analysis and measure theory. We introduce the necessary conceptual framework in order to achieve this objective and then rigorously develop the two formalisms that historically guided the evolution of the study of independence, as well as the surprising parallelism between them. Crossing the border between the population and empirical worlds, we present in a concise way a corpus of results related to U-statistics, a fundamental abstraction of estimation theory. The purpose of this chapter is to provide the machinery needed to give way to the empirical formulation of testing given a sample whether two random variables are independent or not. Finally, the original contributions are condensed in the two following chapters, providing a bridge between what has been exposed up to this point and the most current machine learning theory, in order to show the applicability of this work in the field of materials physics and the novel representation of their structure in the form of graphs
Description
Traballo Fin de Grao en Matemáticas. Curso 2020-2021
Keywords
Bibliographic citation
Relation
Has part
Has version
Is based on
Is part of
Is referenced by
Is version of
Requires
Sponsors
Rights
Atribución-NoComercial-CompartirIgual 4.0 Internacional



