A kernel method for unsupervised network inference
Network inference is the problem of inferring edges between a set of real-world objects, for instance, between pairs of proteins in bioinformatics. Current kernel-based approaches to this problem all share a set of common features: (i) they are supervised and hence require labeled training data; (ii) edges in the network are treated as mutually independent and hence topological properties are largely ignored; (iii) they lack a statistical interpretation. We argue that these common assumptions are often undesirable for network inference, and propose (i) an unsupervised kernel method (ii) that takes the global structure into account and (iii) is statistically motivated. Furthermore, it allows us to statistically explain common heuristics applied to network inference and to improve unsupervised protein interaction prediction.