Theoretical Statistics and Mathematics Unit, ISI Delhi

August 23, 2017 (Wednesday) ,
3:30 PM at Webinar

Speaker:
Ghurumuruhan Ganesan,
NYU AbuDhabi

Title:
Phase Transition in Inhomogenous Erdos Renyi random graphs

Abstract of Talk

Consider the complete graph on~\(n\) vertices where edge~\(e\) is
open with probability \(p_n(e)\) independent of other edges. If~\(p_n(e) =
\frac{C}{n}\) is the same for all edges, then the resulting random graph is
homogenous and phase transition here is well studied using a combination of
branching processes and random walks arguments. In this talk, we first
demonstrate that the above analysis is not directly applicable even under
slight inhomogeneity of the edge probabilities. We then use a tree counting
argument to establish the existence of phase transition and determine the
corresponding critical value, for a wide class of inhomogenous random
graphs.