1 paper
Mehmet Pilanci, Elif Vural
A common assumption in semi-supervised learning with graph models is that the class label function varies smoothly on the data graph, resulting in the rather strict prior that the…