Publication:
Goodness-of-fit testing in two dimensions
Goodness-of-fit testing in two dimensions
dc.contributor.advisor | Rolke, Wolfgang A. | |
dc.contributor.author | Bussing, Anderson | |
dc.contributor.college | College of Arts and Sciences - Sciences | en_US |
dc.contributor.committee | Santana Morant, Dámaris | |
dc.contributor.committee | Lorenzo González, Edgardo | |
dc.contributor.department | Department of Mathematics | en_US |
dc.contributor.representative | Cabrera Rios, Mauricio | |
dc.date.accessioned | 2022-04-11T21:19:38Z | |
dc.date.available | 2022-04-11T21:19:38Z | |
dc.date.issued | 2022-03-02 | |
dc.description.abstract | A goodness-of-fit (GOF) test is a test used to check whether a sample of data came from a specific probability density. For example, perhaps a particular sample of data looks be evenly distributed across its domain, and one wishes to test if the sample came from a uniform density or not. While many GOF methods are known for univariate data, much less work has been done on multivariate data. In this thesis we will focus specifically on the two-dimensional (bivariate) case. We will provide background on some of the most popular methods, including the famous Chi-square and Kolmogorov-Smirnov tests. We will detail how to calculate the necessary test statistics, and we will provide a variety of power studies to demonstrate each method’s effectiveness in different scenarios. Lastly, we will implement a method of adjusted p-values, where several different goodness-of-fit tests are combined in a particular way to yield a test with high power across all scenarios. | en_US |
dc.description.abstract | Una prueba de bondad de ajuste es una preuba usada para probar si una muestra de datos vino de una cierta densidad de probabilidad. Por ejemplo, puede que una muestra de datos parezca ser distribuida uniformamente sobre todo su dominio, y se quiere probar si dicha muestra vino de una densidad uniforme o no. Mientras existen muchos m etodos de bondad de ajuste para datos univariados, se han recibido mucho menos enfoque los m etodos para datos multivariados. En esta tesis, enfocaremos espec camente en el caso de datos con dos dimensiones. Proveeremos trasfondo para algunos de los m etodos m as populares, incluyendo las famosas pruebas Chi-square y Kolmogorov-Smirnov. Daremos detalles sobre c omo se calculan los estad sticos de prueba, y realizaremos una variedad de estudios de potencia estad stica para demostrar la e ciencia de cada m etodo en varios escenarios. Por ultimo, implementaremos un m etodo de valores p ajustados, donde se combinan varias pruebas diferentes de una manera particular para dar una prueba con buena potencia estad stica en todos los escenarios. | en_US |
dc.description.graduationSemester | Spring | en_US |
dc.description.graduationYear | 2022 | en_US |
dc.identifier.uri | https://hdl.handle.net/20.500.11801/2870 | |
dc.language.iso | en | en_US |
dc.rights.holder | (c) 2022 Anderson Bussing | en_US |
dc.subject | Kolmogorov | en_US |
dc.subject | Chi Squared | en_US |
dc.subject | Anderson-Darling | en_US |
dc.subject | Characteristic Function | en_US |
dc.subject | p-value | en_US |
dc.subject.lcsh | Goodness-of-fit tests | en_US |
dc.subject.lcsh | Statistical hypothesis testing | en_US |
dc.subject.lcsh | Statistical hypothesis testing--Computer programs. | en_US |
dc.subject.lcsh | Chi-square test | en_US |
dc.title | Goodness-of-fit testing in two dimensions | en_US |
dc.type | Thesis | en_US |
dspace.entity.type | Publication | |
thesis.degree.discipline | Mathematical Statistics | en_US |
thesis.degree.level | M.S. | en_US |
Files
Original bundle
1 - 1 of 1
No Thumbnail Available
- Name:
- ESTA_BussingA_2022.pdf
- Size:
- 1.3 MB
- Format:
- Adobe Portable Document Format
- Description:
License bundle
1 - 1 of 1
No Thumbnail Available
- Name:
- license.txt
- Size:
- 2.26 KB
- Format:
- Item-specific license agreed upon to submission
- Description: