Search In this Thesis
   Search In this Thesis  
العنوان
Statistical Methods for Multi-type Recurrent Event Data Based on Monte Carlo EM Algorithms and Copula Frailties /
المؤلف
Bedair, Khaled Farag.
هيئة الاعداد
باحث / خالد فرج بدير
مشرف / يلى هونج
مشرف / انيونج كيم
مشرف / ايريك سميث
الموضوع
Statistics.
تاريخ النشر
2014.
عدد الصفحات
p 114. :
اللغة
الإنجليزية
الدرجة
الدكتوراه
التخصص
الإحصاء والاحتمالات
تاريخ الإجازة
29/12/2014
مكان الإجازة
جامعة طنطا - كلية التجارة - Statistics
الفهرس
Only 14 pages are availabe for public view

from 127

from 127

Abstract

Many studies presented a variety of intensity models for the single-type with or without informative censoring which are depend on different approaches to model the baseline intensity and frailty terms as well as different estimation methods. The baseline intensity function is either assumed to be completely unspecified (semiparametric model) or to follow a distribution depending on a low-dimensional unknown parameter vector (e.g., Weibull, Gompertz). Other studies fit the shared frailty model by using splines to model the baseline intensity function. Splines with specified number of knots for the baseline hazard are parametric models. But similar to the models with piecewise constant baseline intensity, it is much more flexible compared to the classical parametric models assuming a Gompertz or Weibull distribution. If the number of pieces becomes large, the models show similar flexibility as semiparametric models. In many situations, it is reasonable to expect fitting parametric baseline intensity is unrealistic.