MMTM (Mixed Membership Triangular Model) Qirong Ho, Junming Yin, Eric P. Xing School of Computer Science Carnegie Mellon University
MMTM is a scalable mixed-membership network community detection algorithm, built upon a triangular representation of the network, rather than the typical edge-based or adjacency-matrix-based representations. By using this triangular representation and principled subsampling, MMTM runs in O(N) linear time, where N is the number of network nodes. The current implementation of MMTM uses Collapsed Gibbs Sampling, based on the following paper: On Triangular versus Edge Representations - Towards Scalable Modeling of Networks. Qirong Ho, Junming Yin and Eric P. Xing. NIPS 2012.