Learning Spectral Graph Transformations for Link Prediction

We present a unified framework for learning link prediction and edge weight prediction functions in large networks, based on the transformation of a graph’s algebraic spectrum. Our approach generaliz