Erdős–Rényi model

Erdős–Rényi model
Name	Erdős–Rényi model
Field	Graph theory
Introduced	1959–1960
Creators	Paul Erdős; Alfréd Rényi

Erdős–Rényi model is a foundational random graph model introduced by Paul Erdős and Alfréd Rényi that formalizes the study of probabilistic properties of networks in the tradition of Andrey Kolmogorov's probability theory and John von Neumann's applied mathematics. It provided rigorous results that influenced subsequent work by Paul Turán, Alfréd Rényi's contemporaries, and later researchers such as Béla Bollobás, Fan Chung, and László Lovász. The model connects to classical results in André Weil's combinatorics, Erdős–Ko–Rado theory, and methods used in Stanislaw Ulam's problems.

Definition and variants

The original construction by Paul Erdős and Alfréd Rényi appears alongside contemporaneous treatments by Frank Harary and later expositions by Béla Bollobás: two canonical variants are parameterized by node count and edge probability, reflecting influences from Jakob Bernoulli's binomial law and Srinivasa Ramanujan's asymptotic methods. The G(n,p) variant selects each potential edge independently with probability p, invoking techniques used by Andrey Kolmogorov and Kolmogorov complexity-adjacent probability, while G(n,M) conditions on exactly M edges, echoing combinatorial enumeration methods from G. H. Hardy and J. E. Littlewood. Extensions include directed versions influenced by Norbert Wiener's stochastic ideas, labeled and unlabeled formulations linked to Arthur Cayley's counting, and bipartite adaptations related to problems studied by Leonhard Euler and Srinivasa Ramanujan.

Mathematical properties

Erdős and Rényi derived threshold functions and asymptotic probabilities using combinatorial and probabilistic tools from Paul Erdős's prolific toolkit, integrating concentration inequalities reminiscent of work by Sergei Bernstein and limit theorems connected to Andrey Kolmogorov and Paul Lévy. Key results employ generating functions in the style of G. H. Hardy and analytic combinatorics later popularized by Philippe Flajolet and Robert Sedgewick. Limit distributions relate to classical theorems of Aleksandr Lyapunov and Pafnuty Chebyshev, while structural counting arguments use extremal techniques from Paul Turán and Erdős–Ko–Rado frameworks.

Phase transitions and giant component

Erdős–Rényi graphs exhibit sharp phase transitions first identified in papers by Paul Erdős and Alfréd Rényi, paralleling phenomena in Ludwig Boltzmann's statistical mechanics and criticality studied by Lev Landau. Around average degree one, a unique giant component emerges, a concept later connected to percolation theory advanced by Broadbent and Hammersley and statistical models from Kenneth Wilson and Michael Fisher. Rigorous proofs draw on branching processes related to work by Andrei Kolmogorov and Otto Haldane, while finite-size scaling analogies reference methods of Kenneth Wilson and Leo Kadanoff.

Degree distribution and clustering

In G(n,p) the degree distribution converges to a binomial law approximated by a Srinivasa Ramanujan-style Poisson limit for sparse regimes, echoing limit relations explored by Jakob Bernoulli and refined by Siméon Denis Poisson. Clustering coefficients in Erdős–Rényi graphs remain low compared with models inspired by Duncan Watts and Albert-Barabási scale-free frameworks; comparisons invoke studies by Duncan Watts, Steven Strogatz, and Albert-László Barabási. Analytic techniques use moment methods reminiscent of Kolmogorov-type inequalities and martingale approaches employed by Joseph Doob.

Connectivity and component structure

Classical theorems by Paul Erdős and Alfréd Rényi identify thresholds for connectivity analogous to phase boundaries in work by John von Neumann and Richard Feynman's probabilistic intuition. The size distribution of small components and the absence or presence of isolated vertices relate to occupancy problems treated by George Pólya and enumeration methods of Arthur Cayley, while sharp threshold proofs utilize techniques later systematized by Ehud Friedgut and Jeff Kahn.

Algorithms and generation methods

Efficient random graph generation algorithms for the Erdős–Rényi ensembles build on uniform sampling ideas from Donald Knuth and rejection methods with roots in John von Neumann's random number generation. Implementations exploit pseudo-random generators originally developed in contexts by John von Neumann and later practical software influenced by Dennis Ritchie and Ken Thompson. Analytical sampling complexity connects to algorithmic graph theory advances by Robert Tarjan and Niklaus Wirth.

Applications and extensions

The Erdős–Rényi model underpins probabilistic reasoning in fields influenced by figures such as Claude Shannon in information theory, Norbert Wiener in cybernetics, and Alan Turing in computation. It serves as a null model in empirical studies by Duncan Watts and Steven Strogatz and as a baseline against which Albert-László Barabási's complex network models are contrasted. Extensions include inhomogeneous random graphs related to work by Remco van der Hofstad, geometric random graphs referencing Paul Erdős's collaborations with Miklós Simonovits, and random hypergraphs building on generalizations studied by Béla Bollobás and Vladimir Voevodsky's combinatorial contemporaries.

Category:Random graphs