High-dimensional variable selection for GLMs and survival models

Pazira, H., 2017, [Groningen]: University of Groningen. 177 p.

Research output: ThesisThesis fully internal (DIV)Academic

Copy link to clipboard


  • Hassan Pazira
The focus of the thesis is on the statistical numerical approaches to fit sparse genomic data with GLM and survival data. The thesis describes on the selection of explanatory variables that may affect a univariate outcome. The outcome has a probability distribution that falls in the class of the exponential dispersion family. The approach that is explored is the differential geometry least angle regression (dgLARS) that is developed for generalized linear models. The dgLARS approach is compared to alternative methods for variable selection in generalized linear models. The numerical procedures of dgLARS is improved for the general setting, and is referred to as the extended dgLARS. Moreover, we investigate how well the dispersion parameter of the family of exponential distributions can be estimated. In the meantime, we focus on survival data and the genomic influence, using the relative risk function. In all chapters it is shown that the improved and developed numerical procedures is fast and accurate in the estimation of parameters. In the end, a full description of the code{R} package that has been developed to do all the analysis is presented.
Original languageEnglish
QualificationDoctor of Philosophy
Awarding Institution
  • Wit, Ernst, Supervisor
  • van den Heuvel, E. R., Assessment committee, External person
  • zu Eulenberg, Christine , Assessment committee, External person
  • Mineo, Angelo M., Assessment committee, External person
Award date10-Jul-2017
Place of Publication[Groningen]
Print ISBNs978-90-367-9953-9
Electronic ISBNs978-90-367-9952-2
Publication statusPublished - 2017

Download statistics

No data available

ID: 44052987