Semidefinite Programming for Community Detection With Side Information

This paper produces an efficient semidefinite programming (SDP) solution for community detection that incorporates non-graph data, which in this context is known as side information. SDP standard on graphs. We formulate a semi-definite relaxation the maximum likelihood estimation of node labels, subject to observing both graph and data. formulation distinct from detection, but maintains its desirable properties. calculate exact recovery threshold three types information, are called information: partially revealed noisy well multiple observations (features) per with arbitrary finite cardinality. find has same presence information Thus, methods developed herein computationally asymptotically accurate Simulations show asymptotic results can also shed light performance graphs modest size.