Statistical inference

The research has a focus on problems, which are or have an expectation to be typical, at the Faculty of Forestry. This leads to a concentration on spatial statistics, general methods for statistical inference and sampling methods. It is necessary to develop and modify methods and models under so-called non-standard conditions under which many statistical methods do not work properly, but conditions which appear frequently in forestry applications. As an example methods have do be developed for non-stationary models in space and time. Two applications are classification in image analysis using the wavelet transform and quality assessment of thematic maps using sophisticated resampling methods. Other important research areas are general estimation methods, which works properly also when classical methods fail and nearest-neighbour techniques applied to a variety of problems.