CMStatistics 2022: Start Registration
View Submission - CMStatistics
Title: Some undirected graphical models for circular variables Authors:  Agnese Panzera - University of Florence (Italy) [presenting]
Anna Gottard - University of Firenze (Italy)
Abstract: Graphical models are a powerful probabilistic tool for studying the conditional independence structure of a set of random variables. This class of multivariate models expresses conditional independence by missing edges in a graph. The associated graph is undirected when all the variables are on equal footing. Then the model is called an undirected graphical model. We explore some multivariate circular distributions focusing on their properties about conditional independence and examine the corresponding classes of undirected graphical models. We discuss some related issues and present some applications in protein folding understanding.