An Unbiased Symmetric Matrix Estimator for Topology Inference Under Partial Observability

Network topology inference is a fundamental problem in many applications of network science, such as locating the source fake news, brain connectivity networks detection, etc. Many real-world situations suffer from critical that only limited part observed nodes are available. This letter considers under framework partial observability. Based on vector autoregressive model, we propose novel unbiased estimator for symmetric with Gaussian noise and Laplacian combination rule. Theoretically, prove it converges to matrix probability. Furthermore, by utilizing mixture model algorithm, an effective algorithm called Gauss developed infer structure. Finally, compared state-of-the-art methods, numerical experiments demonstrate proposed enjoys better performance case small sample sizes.