Contrastes de bondad de ajuste
Loading...
Identifiers
Publication date
Authors
Advisors
Tutors
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
El desarrollo de este trabajo se centra en el estudio teórico y práctico del test ji−cuadrado, el cual se utiliza para contrastar cierta hipótesis sobre una muestra de datos. Se presenta, en primer lugar, el contexto histórico del test, incluyendo su origen y evolución. En el estudio teórico, se formaliza el test, con la definición de su respectivo estadístico de contraste y con los principales resultados relacionados con el mismo, incluyendo aquellos sobre la convergencia asintótica del estadístico. Se diferencian el caso en el que la hipótesis nula es simple y el caso en el que se trata de una familia paramétrica. En el ámbito práctico, se realizan simulaciones del test en sus distintos casos mediante la herramienta R, con la finalidad de analizar y caracterizar su comportamiento real. Se estudian el calibrado del test (casos en los que se cumple la hipótesis nula) y su potencia (casos en los que no se cumple la hipótesis nula). También se ponen a prueba ciertas recomendaciones prácticas mencionadas durante el grado.
The development of this project focuses on the theoretical and practical study of the chisquared test, which is used to test certain hypotheses on a data sample. Firstly, the historical context of the test is presented, including its origin and evolution. In the theoretical study, the test is formalized by defining its corresponding test statistic and presenting the main results related to it, including those concerning the asymptotic convergence of the statistic. The case in which the null hypothesis is simple is distinguished from the case in which it belongs to a parametric family. In the practical part, simulations of the test are carried out in its different cases using the R software, with the aim of analyzing and characterizing its actual behavior. The calibration of the test (cases in which the null hypothesis holds) and its power (cases in which the null hypothesis does not hold) are studied. Certain practical recommendations mentioned during the degree are also tested.
The development of this project focuses on the theoretical and practical study of the chisquared test, which is used to test certain hypotheses on a data sample. Firstly, the historical context of the test is presented, including its origin and evolution. In the theoretical study, the test is formalized by defining its corresponding test statistic and presenting the main results related to it, including those concerning the asymptotic convergence of the statistic. The case in which the null hypothesis is simple is distinguished from the case in which it belongs to a parametric family. In the practical part, simulations of the test are carried out in its different cases using the R software, with the aim of analyzing and characterizing its actual behavior. The calibration of the test (cases in which the null hypothesis holds) and its power (cases in which the null hypothesis does not hold) are studied. Certain practical recommendations mentioned during the degree are also tested.
Description
44 páxs
Keywords
Bibliographic citation
Relation
Has part
Has version
Is based on
Is part of
Is referenced by
Is version of
Requires
Sponsors
Rights
Attribution-NonCommercial-ShareAlike 4.0 International







