To content
Department of Statistics

Topics for Theses

This page lists potential topics for theses in different research areas.

Survival Time Analysis

  • Inference and uncertainty quantification for multi-stage models
  • Comparisons based on reconstructed data
  • Estimation procedures in survival time analysis using machine learning techniques
  • Comparative simulation studies
  • Quantile-based ANOVA methods in complex designs
  • Evaluation possibilities of diagnostic tests using ROC curves and partial AUC

Machine learning and prediction models

  • Influence of missing values on prediction and inference quality of statistical learning methods
  • Uncertainty quantification in the form of variance estimators, prediction and confidence intervals based on machine learning methods
  • Variable selection and dimension reduction based on machine learning methods
  • Time series analysis based on machine learning methods

Analysis of multivariate data

Multivariate Testing Procedures and Multiple Testing

  • Quadratic-form-based multiple contrast tests 
  • Multivariate Non-Inferiority und gerichtete Hypothesen 
  • Einfluss der Wahl einer Hypothesenmatrix

Covariate-Adjusted Statistical Methods

  • common and hypothesis-specific adjustments 
  • selection and optimization of covariate adjustment 
  • simulation-based comparisons of different NANCOVA approaches

Covariance, Correlation, and Reliability Measures

  • Cronbach's alpha and related reliability measures 
  • functionals of covariance and correlation matrices 
  • methods for incomplete or complex multivariate data

Confidence Regions and Finite-Sample Inference

  • comparison of different confidence regions
  • asymptotic and resampling-based methods

Statistical Software and Efficient Implementation

  • implementation and extension of statistical methods in R 
  • efficient algorithms for multivariate and high-dimensional methods

TeX Template

You can use the following TeX template for your thesis: Download


Contact