Estimating Random walk Centrality

Loading...
Thumbnail Image
Date
2019
Authors
Epasinghege Dona, Nirodha Mihirani
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Centrality measures play an important role in determining the importance of vertices in networks. For strongly connected networks, the random walk centrality measures how easy it is to reach a given state from another randomly chosen state. This measure requires calculating a generalized group inverse for the transition matrix, which can be computationally difficult for large state spaces. It is known that the random walk centrality for a particular state can be written as a function of the first and second moments of the inter-arrival times for that state. In this study, using the realization of random walks, we estimate these moments by using a number of statistical methods, including Bayesian bootstrap and two Poisson mixture model approaches. Finally, we compare the resulting estimates of the random walk centrality measures to the true values.
Description
Keywords
Random Walk Centrality, Bayesian bootstrap, Finite Poisson Mixtures, Bayesian Analysis
Citation