[Ml-stat-talks] Scalable Latent Variable Modeling in the Social Sciences

David Blei blei at CS.Princeton.EDU
Wed Jan 2 22:02:02 EST 2013

hi ml-stat-talks,

sean gerrish will present his dissertation on tuesday january 8 at
1:00pm in room 402 of the CS building.  please come if you are
interested in scalable latent variable modeling for the social
sciences (especially political science), probabilistic graphical
models, and/or approximate posterior inference with variational

details below.



Applications of Latent Variable Models in Modeling Influence and Decision Making
Sean Gerrish

Tuesday January 8, 2013 at 1:00PM
Computer Science Building Room 402


The past several decades have seen an explosion of digitized information
in fields such as political science, academics, and news media.  In this
period, a variety of statistical tools have been developed to better
understand patterns in these records.  In this talk, I will present several
recent statistical models for discovering patterns in the fields of politics,
bibliometrics, and international relations. I will describe three models
emphasizing the significance of text data. One of these models uses the
discourse of academic journals to predict the influence of documents.
Another uses the text of legislative documents to understand the issues
driving lawmakers' votes.  A final model uses newspaper articles to model
countries' relations with one another.  I will briefly touch upon details of
approximate posterior inference for these models.

More information about the Ml-stat-talks mailing list