El coeficiente de correlación: desde la independencia lineal a la independencia general de variables aleatorias
Loading...
Identifiers
Publication date
Authors
Advisors
Tutors
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
[ES] En este trabajo haremos un repaso al concepto de coeficiente de correlación. Empezaremos viendo el coeficiente de correlación de Pearson y sus propiedades. Tras estudiar este coeficiente, veremos que sus limitaciones nos llevan a buscar nuevos coeficientes que sean capaces de superarlas. Así surgirá el coeficiente de correlación de Spearman, que nos lleva al estudio de los rangos para descubrir las relaciones de dependencia entre variables. Posteriormente, el siguiente coeficiente, la τ de Kendall, tendrá una gran cantidad de variantes que le permitirá adaptarse a las necesidades que tengamos. Destacamos también la capacidad de la τ de Kendall de adaptar el concepto del coeficiente de correlación de rangos a condiciones poblacionales. El último caso de coeficientes de correlación que veremos es el coeficiente de correlación de distancias. Este último, que nace para solucionar los problemas existentes en los casos multidimensionales, es el más reciente y nos obligará a hacer una breve revisión a la teoría. Finalmente, daremos unas nociones de las funciones cópula. Veremos sus propiedades y definiremos grupos de las funciones cópula arquimedianas y las funciones cópula gaussianas. Si bien las funciones cópula no están directamente influenciadas por la correlación, existen diversas relaciones, principalmente a través de las cópulas gaussianas. Acabaremos con unas breves observaciones realizadas sobre el uso de estas funciones en diversos ámbitos
[EN] Our intention in this essay is to review the concept of the correlation coeficient. We will start by taking a look on Pearson’s correlation coeficient and it’s properties. After that, the own coeficient’s limitations will force us to keep searching new coeficients that are able to overcome this limitations. By this search the Spearman’s correlation coefficient will appear, guiding us to discover dependence relations between variables by the study of ranks. Following, we will review the next coeficient, Kendall’s τ , which will have a large number of variations that will allow it’s adaptation to our own needs. Will be worth to remark that the Kendall’s τ is able to applay the rank’s correlation coefficient concept to population conditions. The last case of correlation coefficients that will be studied will be the distance correlation coeficient. This last one, which it’s defined to solve the problems that exist in multidimensional situations, it’s the most recent of all and will force us to make a slight review of it’s theory. To give the essay an end, we will give some notions about the copula functions. We will see it’s properties and define the groups of arquimedean copula functions and gaussian copula functions. Although copula functions are not directly infuenced by correlation, there are several relations, mostly through gaussian copulas. We will finish by giving some short observations about the use of this functions
[EN] Our intention in this essay is to review the concept of the correlation coeficient. We will start by taking a look on Pearson’s correlation coeficient and it’s properties. After that, the own coeficient’s limitations will force us to keep searching new coeficients that are able to overcome this limitations. By this search the Spearman’s correlation coefficient will appear, guiding us to discover dependence relations between variables by the study of ranks. Following, we will review the next coeficient, Kendall’s τ , which will have a large number of variations that will allow it’s adaptation to our own needs. Will be worth to remark that the Kendall’s τ is able to applay the rank’s correlation coefficient concept to population conditions. The last case of correlation coefficients that will be studied will be the distance correlation coeficient. This last one, which it’s defined to solve the problems that exist in multidimensional situations, it’s the most recent of all and will force us to make a slight review of it’s theory. To give the essay an end, we will give some notions about the copula functions. We will see it’s properties and define the groups of arquimedean copula functions and gaussian copula functions. Although copula functions are not directly infuenced by correlation, there are several relations, mostly through gaussian copulas. We will finish by giving some short observations about the use of this functions
Description
Traballo Fin de Grao en Matemáticas. Curso 2019-2020
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



