According to the authors of recent research from Santa Clara, California, "Co-training (or more generally,nco-regularization) has been a popular algorithm for semi-supervised learning in data with two featurenrepresentations (or views), but the fundamental assumptions underlying this type of models are stillnunclear. In this paper we propose a Bayesian undirected graphical model for co-training, or morengenerally for semi-supervised multi-view learning."
展开▼