Abschlussarbeiten
Aktuell offene Themen
Master
- Spatio-temporal modelling of effects on rental vehicles usage using open data
Several rental bike or e-scooter providers make APIs available to retrieve the locations of their verhicles. The aim of this work is to analyse the usage of rental verhicles using spatio-temporal models and additional open data like Open Street Map, weather data from DWD etc.
- Nested Sampling for Geoadditive Models
Nested Sampling (Skilling, Bayesian Analysis 2006) is frequently used in astronomy for sampling posteriors and for model comparison. However, it has not applied in classical spatial statistics applications. The aim of this work is to adapt the algorithm to spatial models with latent GMRF (or GRF). Implementation could be done in Julia, R, C(++) or python. The application of the algorithm should than be compared to several MCMC methods (and maybe approximation methods) in respect to computational resources.
- Three-level Regional Adaptive Smoothing for Disease Mapping
Smoothing is a usefull technique in disease mapping to "borrow strength" from neighbouring administrative areas. However, smoothing using a global smoothing paramter is often not appropriate. Building on an existing two-level model (Ali 2021), the aim of this thesis is to develop and apply a spatial model with regional smoothing parameters for data on at least three levels of administrative areas (e.g., Bundesland, Landkreis, Gemeinde).
- Colocalization analysis in Flourescence Microscopy using ideas from Spatial Statistics
One of the aims of flourescence microscopy is to determine the colocalization of molecules in the cell nucleus. The aim of this thesis is to model the colocalization using ideas from spatial point process regression, without the use of the computational expensive spatially structured components. The developed method should be applied to a simulated and one more multi-image data sets.