Posts

Buffered target-weighted cross-validation (TWCV) aligns validation with deployment by reweighting prediction tasks based on prediction-distance and covariate distributions.

Day by day, 24 concepts from geographic information science that every geographer should know…

A new version of RSAGA is available, the R package that provides an interface to SAGA GIS for geospatial analysis, digital terrain analysis, and geocomputing…

Summary of my expert report on regionalization methods for the designation of areas with nitrate-polluted groundwater, prepared for the German federal environmental agency…

Machine-learning methods for natural hazards assessment are data hungry — in her PhD, Zhihao Wang developed an active–transfer learning strategy that combines transferred prior knowledge with optimal local data acquisition…

The Climate Walk is a self-guided tour through the city of Jena, offering insights into urban climate and adaptation strategies…

GIScience group members are contributing to the EGU conference, on-site as well as online. See what we’ve been up to…

Are intelligent chatbots capable of thinking geographically? What do they know about all things local? See how GPT-3 did in my little benchmarking experiment…

How do the predictions of machine-learning and geostatistical models deteriorate with increasing distance to the training locations? Are different predictors instrumental in making predictions at greater versus shorter distances? A novel set of diagnostic tools allows us to answer these questions…

Slope aspect is a predictor in many environmental machine-learning models. Here I compare strategies for dealing with problems caused by the cyclic nature of this directional variable.