• Corpus ID: 6418786

Discovering Latent Network Structure in Point Process Data

  title={Discovering Latent Network Structure in Point Process Data},
  author={Scott W. Linderman and Ryan P. Adams},
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… 

Figures from this paper

Semi-parametric Network Structure Discovery Models
We propose a network structure discovery model for continuous observations that generalizes linear causal models by incorporating a Gaussian process (GP) prior on a network-independent component, and
Mutually exciting point process graphs for modelling dynamic networks
MEG is a scalable network-wide statistical model for point processes with dyadic marks, which can be used for anomaly detection when assessing the significance future events, including previously unobserved connections.
Scalable Bayesian Inference for Excitatory Point Process Networks
This work focuses on the problem of latent network discovery in the case where the observable activity takes the form of a mutually-excitatory point process known as a Hawkes process, and builds on previous work that has taken a Bayesian approach to this problem, specifying prior distributions over the latent network structure and a likelihood of observed activity given this network.
Inferring Multiplex Diffusion Network via Multivariate Marked Hawkes Process
The Multiplex Diffusion Model (MDM) is proposed which incorporates the multivariate marked Hawkes process and topic model and a MCMC based algorithm is developed to infer the latent multiplex structure and to estimate the node-related parameters.
Network Estimation From Point Process Data
This paper incorporates saturation in a point process model which both ensures stability and models non-linear thresholding effects and imposes general low-dimensional structural assumptions that include sparsity, groupSparsity, and low-rankness that allows bounds to be developed in the high-dimensional setting.
Multivariate Spatiotemporal Hawkes Processes and Network Reconstruction
This paper develops a nonparametric method for network reconstruction from spatiotemporal data sets using multivariate Hawkes processes that yields improved network reconstruction, providing a basis for meaningful subsequent analysis of the reconstructed networks.
Network estimation via poisson autoregressive models
  • B. Mark, G. Raskutti, R. Willett
  • Computer Science
    2017 IEEE 7th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP)
  • 2017
This paper provides learning rates for a class of multivariate self-exciting Poisson autoregressive models and shows how the derived learning rates apply in the high-dimensional setting when the network is sparse.
Bayesian Inference of Network Structure From Information Cascades
A probabilistic framework using samples from the distribution of networks that are compatible with the dynamics observed to produce network and uncertainty estimates is presented and the benefits of quantifying uncertainty to improve estimates and understanding are shown, particularly with small amounts of data.
Tracking Dynamic Point Processes on Networks
The proposed approach is an online learning framework well-suited to streaming data, using a multivariate Hawkes point process model to encapsulate autoregressive features of observed events within the social network.
Causal Discovery in High-Dimensional Point Process Networks with Hidden Nodes
This work proposes a deconfounding procedure to estimate high-dimensional point process networks with only a subset of the nodes being observed, which allows flexible connections between the observed and unobserved processes.


Latent Point Process Models for Spatial-Temporal Networks
This work develops an efficient approximate algorithm based on variational expectationmaximization to infer unknown participants in an event given the location and the time of the event.
Point process modelling for directed interaction networks
A model is introduced for treating directed interactions as a multivariate point process: a Cox multiplicative intensity model using covariates that depend on the history of the process, and consistency and asymptotic normality are proved.
Learning Social Infectivity in Sparse Low-rank Networks Using Multi-dimensional Hawkes Processes
This paper proposes a convex optimization approach to discover the hidden network of social influence by modeling the recurrent events at different individuals as multidimensional Hawkes processes, emphasizing the mutual-excitation nature of the dynamics of event occurrence.
Modelling Reciprocating Relationships with Hawkes Processes
A Bayesian nonparametric model is presented that discovers implicit social structure from interaction time-series data and outperforms general, unstructured Hawkes processes as well as structured Poisson process-based models at predicting verbal and email turn-taking, and military conflicts among nations.
Random function priors for exchangeable arrays with applications to graphs and relational data
A flexible yet simple Bayesian nonparametric model is obtained by placing a Gaussian process prior on the parameter function which constitutes the natural model parameter in a Bayesian model.
Inferring Networks of Diffusion and Influence
This work develops an efficient approximation algorithm that scales to large datasets and finds provably near-optimal networks for tracing paths of diffusion and influence through networks and inferring the networks over which contagions propagate.
Modeling homophily and stochastic equivalence in symmetric relational data
A latent variable model for inference and prediction of symmetric relational data, based on the idea of the eigenvalue decomposition, that generalizes other popular latent variable models.
A Survey of Statistical Network Models
An overview of the historical development of statistical network modeling is overviewed and a number of examples that have been studied in the network literature are introduced, and a subsequent discussion focuses on anumber of prominent static and dynamic network models and their interconnections.
Discovering latent influence in online social activities via shared cascade poisson processes
The proposed probabilistic model for discovering latent influence from sequences of item adoption events based on the stochastic EM algorithm can be used for finding influential users, discovering relations between users and predicting item popularity in the future.
Elliptical slice sampling
A new Markov chain Monte Carlo algorithm for performing inference in models with multivariate Gaussian priors is presented, which has simple, generic code applicable to many models, and works well for a variety of Gaussian process based models.