Local diversity characteristics and hotspot detection in heterogeneous point patterns

Modern technologies enable the collection of extremely large and heterogeneous spatial point pattern datasets, such as mapping all trees in an entire country or all cells in a medical tissue sample. Efficient analysis of such new datasets requires new statistical methods. The aim of the DIVSPOT project is to develop methods for quantifying, estimating, and visualizing local diversity characteristics in large, heterogeneous point patterns, as well as to provide practical tools for discovering and identifying new phenomena.

These methods are needed, for example, in environmental monitoring and cancer research, where it is essential to identify regions in which diversity is particularly high or otherwise significantly different. Current methods are insufficient both computationally and statistically, which reduces the quality of research and limits the effective use of new datasets.

The project builds on the theory of point processes and spatial statistics to develop reliable and efficient approaches for analyzing diversity, detecting hotspot regions, and applying the results to key research questions in ecology and medicine in collaboration with domain experts.