Discovering Latent Network Structure in Point Process Data
Networks play a central role in modern data analysis, enabling us to reason about systems by studying the relationships between their parts. Most often in network analysis, the edges are given. However, in many systems it is difficult or impossible to measure the network directly. Examples of latent networks include economic interactions linking financial instruments and patterns of reciprocity in gang violence. In these cases, we are limited to noisy observations of events associated with each node. To enable analysis of these implicit networks, we develop a probabilistic model that combines mutually-exciting point processes with random graph models. We show how the Poisson superposition principle enables an elegant auxiliary variable formulation and a fully-Bayesian, parallel inference algorithm. We evaluate this new model empirically on several datasets.
Scott Linderman is a Computer Science Ph.D. candidate at Harvard University where he is quite fortunate to be advised by Leslie Valiant and Ryan Adams and to be a member of the Harvard Intelligent Probabilistic Systems group. Prior to beginning graduate school, he worked at Microsoft for three years as a software engineer on the Windows networking stack. His research focuses on computational neuroscience and machine learning, primarily on probabilistic models and inference algorithms for discovering structure in large-scale neural recordings. Though his work is often motivated by questions in brain science, he is always open to exploring broader scientific applications.
Dana Research Center, 110 Forsyth St., Boston, MA 02115 on the 5th Fl – Large Elevator