Dr. Leyk's research focuses on three thematically overlapping areas of computational GIScience. First, he is interested in the development of geospatial models under uncertainty using probabilistic approaches and fuzzy set theory. He conducts research funded by NSF and NIH on models for small area estimation in demographic and health data. Second, Dr. Leyk's research is on dynamic phenomena in complex integrated systems in space and time. This research focuses on issues of spatial non-stationarity in disease and migration models in order to improve existing spatial statistical models. It also embraces problems due to temporal aggregation and variation in spatial variables to improve the understanding of underlying relationships and their changes over time. Third, Dr. Leyk develops pattern recognition techniques for feature extraction in maps, in particular historical maps, and remotely sensed imagery for land cover change detection and change analysis in population distributions.
uncertainty modeling in GIScience, cartographic pattern recognition, demographic small area estimation, dasymetric modeling, spatio-temporal modeling