Undergraduate Honors Thesis

 

Hidden Erdös-Rényi Graphs in Stochastic Block Model Networks 公开 Deposited

https://scholar.colorado.edu/concern/undergraduate_honors_theses/rx913r321
Abstract
  • In this thesis, we use a new probabilistic and statistical methodology to estimate the fraction of spurious edges in a Stochastic Block model (SBM). By spurious we mean unstructured connections between nodes in the network, which we attribute as coming from an Erdös-Rényi random graph model. We define the “hidden weight” of a SBM as the largest expected fraction of placed or missing edges that can be attributed to an Erdös-Rényi model. We characterize explicitly and asymptotically this weight in terms of the parameters defining the SBM, and test our results with simulations.

Creator
Date Awarded
  • 2022-04-01
Academic Affiliation
Advisor
Granting Institution
最新修改
  • 2022-04-20
Resource Type
权利声明
Language

关联

属于 Collection:

单件