Physics > Physics and Society
[Submitted on 27 Oct 2012 (v1), revised 12 Sep 2013 (this version, v3), latest version 25 Jun 2014 (v4)]
Title:Tractable and Consistent Random Graph Models
View PDFAbstract:We define a general class of network formation models, Statistical Exponential Random Graph Models (SERGMs), that nest standard exponential random graph models (ERGMs) as a special case. We analyze conditions for practical and consistent estimation of the associated network formation parameters, addressing two open issues in the estimation of exponential random graph models. First, there are no previous general results on whether estimates of such a model's parameters based a single network are consistent (i.e., become accurate as the number of nodes grows). Second, a recent literature has shown that standard techniques of estimating ERGMs have exponentially slow mixing times for many specifications in which case the software used for estimating these models will be unreliable. SERGMs reformulate network formation as a distribution over the space of sufficient statistics instead of the space of networks, greatly reducing the size of the space of estimation and making estimation practical and easy. We identify general classes of models for which maximum likelihood estimates are consistent and asymptotically normally distributed. We also develop a related, but distinct, class of models that we call subgraph generation models (SUGMs) that are useful for modeling sparse networks and whose parameter estimates are also consistent and asymptotically normally distributed. We show how choice-based (strategic) network formation models can be written as SERGMs and SUGMs, and illustrate the application of our models and techniques with network data from villages in Karnataka, India.
Submission history
From: Arun Chandrasekhar [view email][v1] Sat, 27 Oct 2012 21:29:23 UTC (253 KB)
[v2] Tue, 30 Oct 2012 02:03:10 UTC (254 KB)
[v3] Thu, 12 Sep 2013 19:33:20 UTC (1,606 KB)
[v4] Wed, 25 Jun 2014 07:55:23 UTC (1,638 KB)
Current browse context:
physics.soc-ph
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Connected Papers (What is Connected Papers?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.