Local Weak Convergence Based Analysis of a New Graph Model

M. Moharrami; V. Subramanian; M. Liu; R. Sundaresan

doi:10.1109/ALLERTON.2018.8635966

Back

Conference proceeding

Local Weak Convergence Based Analysis of a New Graph Model

M. Moharrami, V. Subramanian, M. Liu and R. Sundaresan

2018 56th Annual Allerton Conference on Communication, Control, and Computing (Allerton), pp.502-503

10/2018

DOI: 10.1109/ALLERTON.2018.8635966

View Online

Abstract

Different random graph models have been proposed as an attempt to model individuals' behavior. Each of these models proposes a unique way to construct a random graph that covers some properties of the real-world networks. In a recent work [4], the proposed model tries to capture the self-optimizing behavior of the individuals in which the links are made based on the cost/benefit of the connection. In this paper, we analyze the asymptotics of this graph model. We prove the model locally weakly converges [1] to a rooted tree associated with a branching process which we named Erlang Weighted Tree(EWT) and analyze the main properties of the EWT.

Analytical models

Artificial neural networks

Computational modeling

Convergence

Eigenvalues and eigenfunctions

Probability distribution

Random variables

Details

Title: Subtitle: Local Weak Convergence Based Analysis of a New Graph Model
Creators: M. Moharrami - University of Michigan–Ann Arbor
V. Subramanian - University of Michigan–Ann Arbor
M. Liu - University of Michigan–Ann Arbor
R. Sundaresan - Indian Institute of Science Bangalore
Resource Type: Conference proceeding
Publication Details: 2018 56th Annual Allerton Conference on Communication, Control, and Computing (Allerton), pp.502-503
Publisher: IEEE
DOI: 10.1109/ALLERTON.2018.8635966
ISSN: 2474-0195
Language: English
Date published: 10/2018
Academic Unit: Computer Science
Record Identifier: 9984446274202771

Metrics

4 Record Views