A. Adin Urtasun, E. Orozco-Acosta, M. D. Ugarte
Research in spatio-temporal disease mapping has been very fruitful in recent decades and numerous statistical models have been propose to study the geographic distribution of a disease and its evolution in time. However, the scalability of these models has not been studied in depth yet. In this work, we propose a “divide-and-conquer” based methodology as an alternative modelling approach to the commonly used disease mapping models in order to analyze high-dimensional spatio-temporal data. The methods and algorithms proposed in this work are implemented in the open-source R package "bigDM" (https://github.com/spatialstatisticsupna/bigDM), which allows the user to adapt the modelling scheme to their own processing architecture by performing both parallel and/or distributed computation strategies. Currently, we are working on the development of scalable ecological regression models taking into account the spatial and/or spatio-temporal confounding issues between fixed and random effects.
Keywords: Disease mapping, Hierarchical models, Laplace approximations, Massive data, Scalable modelling
Scheduled
GT19 Time Space Statistics
June 7, 2022 12:00 PM
Conference hall