Open Conference Systems, 50th Scientific meeting of the Italian Statistical Society

Font Size: 
Overlapping mixture models for network data (\texttt{manet}) with covariates adjustment
Saverio Ranciati, Giuliano Galimberti, Ernst C. Wit, Veronica Vinciotti

Last modified: 2018-05-24

Abstract


Network data often come in the form of \emph{actor}-\emph{event} information, where two types of nodes comprise the very fabric of the network. Examples of such networks are: people voting in an election, users liking/disliking media content, or, more generally, individuals - \emph{actors} - attending events. Interest lies in discovering communities among these actors, based on their patterns of attendance to the considered events. To achieve this goal, we propose an extension of the model introduced in \cite{ranciati2017identifying}: our contribution injects covariates into the model, leveraging on parsimony for the parameters and giving insights about the influence of such characteristics on the attendances. We assess the performance of our approach in a simulated environment.

References


  1. Frühwirth-Schnatter, S.: Finite Mixture and Markov Switching Models. Springer Science & Business Media, New York (2006)

  2. Holmes,C.C.,Held,L.:Bayesian auxiliary variable models for binary and multinomial regression. Bayesian analysis 1(1), 145–168 (2006)

  3. Kolaczyk, E.D.: Statistical Analysis of Network Data: Methods and Models. Springer, New York (2009)

  4. McCulloch, C.E., Neuhaus, J.M.: Generalized linear mixed models. Wiley Online Library (2001)

  5. Ranciati,S.,Vinciotti,V.,Wit,E.C.:Identifying overlapping terrorist cells from the noordintop actor-event network. arXiv preprint arXiv:1710.10319 (2017)

  6. Wasserman, S., Faust, K.: Social network analysis: Methods and applications, vol. 8. Cam- bridge university press (1994)


Full Text: PDF