Speaker:
Prof. Stéphane Robin, INRA, France
Time：
December 22, 2020, 15:00-17:00 Beijing time
Zoom Meeting ID：872 5682 0078
Zoom Meeting Password: 123456
Abstract: Graphical models provide a powerful framework to analyze the dependency structure relating a set of random variables. Recently, the inference of the structure of a graphical model has received a lot of attention, and one of the main issue is the exploration of the space of possible graphs, which can not be carried in a naive manner because of combinatorial complexity. Spanning trees constitute a subset of graph, which fulfill the popular sparsity assumption. Still, the tree structure is a too restrictive assumption for most applications. However, the Matrix-Tree theorem enables to integrate over the set of all spanning tree at the cost of the calculation of a determinant, which allows considering a mixture of tree-shaped graphical models. In this talk, we will show how mixtures of tree-shapes graphical models can be used to infer the graphical model of a set of variables and the applications in ecology and epidemiology.