“…One method of recovering the parameters is to observe random walks on the network as a weighted graph and try to recover the transition matrix, as the transition probabilities are often directly related to the weights. The inverse problems for random walks were studied in [47,48,76,79], and applications were considered in optical tomography [54,55,56,57,78], network tomography [58,86], electrical resistor networks [39,65,74] and neuroscience [8]. These works use different types of random walk measurements at accessible nodes to recover the transition matrix in different settings, assuming the topology of the network is known.…”