Latent Space Models

From Cohen Courses
Revision as of 16:33, 1 April 2011 by Ssomanch (talk | contribs)
Jump to navigationJump to search

The important idea in the latent space modelling is that the nodes which are associated with p-dimensional Euclidean space, and points that close in the latent space are more probable to form a link in the network. Seminal paper in the latent space modelling is due to Raftery etal [ http://www.stat.washington.edu/raftery/Research/PDF/hoff2002.pdf Latent space approaches to social network analysis] in Journal of American Statistical Association.