Abstract
We consider the problem of simplifying large weighted directed graphs by aggregating nodes and edges. This problem is recast as a clustering/resource allocation problem, and a solution method that incorporates features of the deterministic annealing (DA) algorithm is proposed. The novelty in our method is a quantitive measure of dissimilarity that allows us to compare directed graphs of possibly different sizes (i.e., the original and the aggregated graphs). The approach we propose is insensitive to initial conditions and less likely to converge to poor local minima than Lloyd-type algorithms. We apply our graph-aggregation (clustering) method to Markov chains, where low-order Markov chains that approximate high-order chains are obtained through appropriate aggregation of state transition matrices. We further develop a decentralized computational scheme to improve tractability of the algorithm.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
