VABAR

The objective of our group is the methodological and applied research of Bayesian Statistics, especially in scenarios of epidemiological and environmental type. Our work is fundamentally based on three axes: Hierarchical models in studies with correlated data. Model selection. Computational simulation models. All of them are somewhat mixed in nature, methodological and applied, and in the compatibility and interrelation of many of their knowledge and objectives. The first thematic block is the most extensive and is dedicated to research on models with correlated data associated with structures of a spatial-temporal, longitudinal, survival or non-survival type, and of blood kinship. Methodological research in disease mapping has a long tradition in our team, currently with unbalanced multivariate and spatial-temporal objectives. This block also contains new research proposals dedicated to joint models with longitudinal and survival data, methodology on species distribution, spatial-temporal surveillance of diseases and regression methods for scattered genetic data from genetically isolated populations with known family trees, which will undoubtedly lead us to Big Data. In the area dedicated to the selection of models, marginal, conditioned, and combined measures are studied to quantify the contribution of a potential set of covariates in the explanation of a response of interest, and two new lines have been initiated, with a more applied orientation, which link the subject of variable selection with longitudinal and survival models and related data through structures of consanguinity. Finally, in the block of computational models, the group continues with a line of action dedicated to the calibration of multivariate computational models and the implementation of the results obtained in a computer application and in a new application dedicated to uncertainty modelling in compartmental models.