[Forum SIS] Avviso di Seminario - Modelli Grafici, D'Errico e Nicolussi

[Dottorato]

Nell'ambito del Dottorato in Statistica dell'Università degli Studi di 
Milano-Bicocca,

è organizzato il seminario

*Introduction to Graph Theory and Graphical Models*

*Marco D'Errico* and Federica Nicolussi***

*  Dipartimento di Statistica e Metodi Quantitativi, Università degli 
Studi di Milano-Bicocca
** Dipartimento di Sociologia e Ricerca Sociale, Università degli Studi 
di Milano-Bicocca

in data 17 Dicembre 2013, dalle h12.00 alle h14.00, presso l'aula 
seminari del dipartimento DEMM, secondo piano, Facoltà di Scienze 
Politiche dell'Università degli Studi di Milano, via Conservatorio 7.

La partecipazione è gratuita, ma è gradita segnalazione di interesse per 
motivi organizzativi.

Segue abstract.

In light of the growing literature on graphical models and complex 
networks, graph theory has recently reemerged as one of the essential 
analytical frameworks to model pairwise relationships between a set of 
items (often referred to as 'nodes' or 'vertices') that are connected by 
'links' (or 'edges'). One of the main applications of graph theory are 
Graphical Models, i.e. probabilistic models where graphs are used to 
indicate the conditional dependence relationship between random variables.

In the first part of this seminar, we will introduce some basic notions 
and concepts in graph theory, with a particular emphasis on connectivity 
and centrality measures, and on the underlying mathematical concepts. We 
will then show some examples of real-world networks, where these 
concepts are successfully applied. Lastly, we will introduce the 
analytical background for Graphical Models.

The second part of the seminar will deal with an overview on different 
types of Graphical Models, with special care to the problematic that 
these models meet. Through examples we will explain how to represent 
different kinds of relationships (unidirectional or bidirectional) 
between variables involved in the model. Particular attention will be 
reserved to the connection between the previous Graphical Models and the 
Log-linear models (and their generalization).


-------------- parte successiva --------------
Un allegato HTML è stato rimosso...
URL: <http://www.stat.unipg.it/pipermail/sis/attachments/20131122/7b04d16d/attachment.html>