Stochastic Block Model

Community structure in random graphs — 3 blocks with tunable inter/intra-community edge probabilities

Edges: 0 Modularity Q: - Detectability: -
About: The Stochastic Block Model (SBM) is the canonical model for networks with community structure. Nodes are partitioned into K blocks; edges form independently with probability p_in within blocks and p_out between them. Community detection is possible if and only if (p_in−p_out)² > K(p_in + (K-1)p_out)/N — the Kesten-Stigum threshold (Mossel, Neeman & Sly 2015). Below this threshold, the structure is information-theoretically undetectable.