You can manage bookmarks using lists, please log in to your user account for this.
Media type:
Doctoral Thesis;
E-Book;
Electronic Thesis
Title:
Estimation of adjusted relative risks in log-binomial regression
Contributor:
Bekhit, Adam
[Author]
imprint:
Saarländische Universitäts- und Landesbibliothek, 2021
Language:
English
DOI:
https://doi.org/10.22028/D291-35023
Origination:
Footnote:
Diese Datenquelle enthält auch Bestandsnachweise, die nicht zu einem Volltext führen.
Description:
**Problem statement:** For binary outcome data, the relative risk (RR) is an essential measure of association, which can be estimated directly for prospective studies. Calculating the odds ratio (OR) can overestimate and magnify the risk heavily, especially if dealing with a disease outcome of high incidence. In such cases OR should be avoided and RR can be used. The log-binomial model is a straightforward statistical approach in the case of risk adjustment and estimation, also it is much easier to interpret. However, the log-binomial model might fail or have difficulties maximizing log-likelihood function due to numerical instability, implicit parameter constraints, or naïve starting value, which leads to a dramatic increase of the required number of iterations therefore convergence failure. **Approach:** In this study, a modified Newton-type algorithm was developed for solving the maximum likelihood estimation problem under linear constraints. Moreover, a new system of linear inequality constraints on the number of covariates was imposed. In this approach, the log-likelihood function of the log-binomial regression model is maximized sequentially. The modified-newton-type algorithm proceeds iteratively by generating a sequence of estimates which solves the quadratic sub-problems obtained from a second-order Taylor approximation and converges under the linear inequality constraints. **Monte Carlo Simulation design:** For validation and evaluation purposes, a large full-factorial simulation study was conducted in order to study the behavior of the method "squadP" compared with other methods investigated in this research such as Fisher scoring, EM-type, BFGS, and Nelder-Mead. Assessment of coverage probability, model accuracy, and model bias were the primary objectives, while at the same time allowing for many different scenarios (varying number of events, sample size, and number of covariates). The 12 underlying covariates were generated via copula package in R with a specified correlation structure between all ...