Title: Graphical models for circular data
Authors: Anna Gottard - University of Firenze (Italy) [presenting]
Agnese Panzera - University of Florence (Italy)
Abstract: Graphical models have been successfully used to characterise conditional independence structures among random variables. For instance, in proteomics, we would now like to analyse the structure of proteins in terms of their characterising angles. For this, the first task is to study conditional independence in multivariate distributions on angles. One example is the multivariate von Mises distribution, also known as the multivariate sine distribution. We review the existing literature on graphical models for circular data and propose possible extensions both for model specifications and inference.