ResearchTrend.AI
  • Communities
  • Connect sessions
  • AI calendar
  • Organizations
  • Join Slack
  • Contact Sales
Papers
Communities
Social Events
Terms and Conditions
Pricing
Contact Sales
Parameter LabParameter LabTwitterGitHubLinkedInBlueskyYoutube

© 2025 ResearchTrend.AI, All rights reserved.

  1. Home
  2. Papers
  3. 1511.00546
192
10
v1v2v3 (latest)

An Impossibility Result for Reconstruction in a Degree-Corrected Planted-Partition Model

2 November 2015
Lennart Gulikers
Marc Lelarge
Laurent Massoulié
ArXiv (abs)PDFHTML
Abstract

We consider a Degree-Corrected Planted-Partition model: a random graph on nnn nodes with two asymptotically equal-sized clusters. The model parameters are two constants a,b>0a,b > 0a,b>0 and an i.i.d. sequence of weights (ϕu)u=1n(\phi_u)_{u=1}^n(ϕu​)u=1n​, with finite second moment Φ(2)\Phi^{(2)}Φ(2). Vertices uuu and vvv are joined by an edge with probability ϕuϕvna\frac{\phi_u \phi_v}{n}anϕu​ϕv​​a when they are in the same class and with probability ϕuϕvnb\frac{\phi_u \phi_v}{n}bnϕu​ϕv​​b otherwise. We prove that it is information-theoretically impossible to estimate the spins in a way positively correlated with the true community structure when (a−b)2Φ(2)≤2(a+b)(a-b)^2 \Phi^{(2)} \leq 2(a+b)(a−b)2Φ(2)≤2(a+b). A by-product of our proof is a precise coupling-result for local-neighbourhoods in Degree-Corrected Planted-Partition models, which could be of independent interest.

View on arXiv
Comments on this paper