Problems in generalized linear model selection and predictive evaluation for binary outcomes
This manuscript consists of three papers which formulate novel generalized linear model methodologies. In Chapter 1, we introduce a variant of the traditional concordance statistic that is associated with logistic regression. This adjusted c − statistic as we call it utilizes the differences in pred...
Main Author: | |
---|---|
Other Authors: | |
Format: | Others |
Language: | English |
Published: |
University of Iowa
2015
|
Subjects: | |
Online Access: | https://ir.uiowa.edu/etd/6003 https://ir.uiowa.edu/cgi/viewcontent.cgi?article=7484&context=etd |
id |
ndltd-uiowa.edu-oai-ir.uiowa.edu-etd-7484 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-uiowa.edu-oai-ir.uiowa.edu-etd-74842019-10-13T04:37:27Z Problems in generalized linear model selection and predictive evaluation for binary outcomes Ten Eyck, Patrick This manuscript consists of three papers which formulate novel generalized linear model methodologies. In Chapter 1, we introduce a variant of the traditional concordance statistic that is associated with logistic regression. This adjusted c − statistic as we call it utilizes the differences in predicted probabilities as weights for each event/non- event observation pair. We highlight an extensive comparison of the adjusted and traditional c-statistics using simulations and apply these measures in a modeling application. In Chapter 2, we feature the development and investigation of three model selection criteria based on cross-validatory c-statistics: Model Misspecification Pre- diction Error, Fitting Sample Prediction Error, and Sum of Prediction Errors. We examine the properties of the corresponding selection criteria based on the cross- validatory analogues of the traditional and adjusted c-statistics via simulation and illustrate these criteria in a modeling application. In Chapter 3, we propose and investigate an alternate approach to pseudo- likelihood model selection in the generalized linear mixed model framework. After outlining the problem with the pseudo-likelihood model selection criteria found using the natural approach to generalized linear mixed modeling, we feature an alternate approach, implemented using a SAS macro, that obtains and applies the pseudo-data from the full model for fitting all candidate models. We justify the propriety of the resulting pseudo-likelihood selection criteria using simulations and implement this new method in a modeling application. 2015-12-15T08:00:00Z dissertation application/pdf https://ir.uiowa.edu/etd/6003 https://ir.uiowa.edu/cgi/viewcontent.cgi?article=7484&context=etd Copyright © 2015 Patrick Ten Eyck Theses and Dissertations eng University of IowaCavanaugh, Joseph E. Biostatistics Generalized linear models Model selection Biostatistics |
collection |
NDLTD |
language |
English |
format |
Others
|
sources |
NDLTD |
topic |
Biostatistics Generalized linear models Model selection Biostatistics |
spellingShingle |
Biostatistics Generalized linear models Model selection Biostatistics Ten Eyck, Patrick Problems in generalized linear model selection and predictive evaluation for binary outcomes |
description |
This manuscript consists of three papers which formulate novel generalized linear model methodologies.
In Chapter 1, we introduce a variant of the traditional concordance statistic that is associated with logistic regression. This adjusted c − statistic as we call it utilizes the differences in predicted probabilities as weights for each event/non- event observation pair. We highlight an extensive comparison of the adjusted and traditional c-statistics using simulations and apply these measures in a modeling application.
In Chapter 2, we feature the development and investigation of three model selection criteria based on cross-validatory c-statistics: Model Misspecification Pre- diction Error, Fitting Sample Prediction Error, and Sum of Prediction Errors. We examine the properties of the corresponding selection criteria based on the cross- validatory analogues of the traditional and adjusted c-statistics via simulation and illustrate these criteria in a modeling application.
In Chapter 3, we propose and investigate an alternate approach to pseudo- likelihood model selection in the generalized linear mixed model framework. After outlining the problem with the pseudo-likelihood model selection criteria found using the natural approach to generalized linear mixed modeling, we feature an alternate approach, implemented using a SAS macro, that obtains and applies the pseudo-data from the full model for fitting all candidate models. We justify the propriety of the resulting pseudo-likelihood selection criteria using simulations and implement this new method in a modeling application. |
author2 |
Cavanaugh, Joseph E. |
author_facet |
Cavanaugh, Joseph E. Ten Eyck, Patrick |
author |
Ten Eyck, Patrick |
author_sort |
Ten Eyck, Patrick |
title |
Problems in generalized linear model selection and predictive evaluation for binary outcomes |
title_short |
Problems in generalized linear model selection and predictive evaluation for binary outcomes |
title_full |
Problems in generalized linear model selection and predictive evaluation for binary outcomes |
title_fullStr |
Problems in generalized linear model selection and predictive evaluation for binary outcomes |
title_full_unstemmed |
Problems in generalized linear model selection and predictive evaluation for binary outcomes |
title_sort |
problems in generalized linear model selection and predictive evaluation for binary outcomes |
publisher |
University of Iowa |
publishDate |
2015 |
url |
https://ir.uiowa.edu/etd/6003 https://ir.uiowa.edu/cgi/viewcontent.cgi?article=7484&context=etd |
work_keys_str_mv |
AT teneyckpatrick problemsingeneralizedlinearmodelselectionandpredictiveevaluationforbinaryoutcomes |
_version_ |
1719264624504733696 |