The Statistical Inference for Network Models symposium is a satellite of the NetSci2017 conference, to be held June 20, 2017 in Indianapolis, Indiana, USA. The exact location of the satellite is TBD. It will feature a mix of invited and contributed talks, which you can view on the Symposium Schedule. All attendees of this symposium must be registered for the NetSci2017 conference in order to attend. Please note the important dates and deadlines on the right.
Call for Abstracts (closed)
We invite abstracts of new and/or previously published work for contributed talks to take place at the symposium. We hope for a broad range of topics to be covered, across theory, methodology, and application to empirical network data. Potential topics include:
Abstract submission will be handled by EasyChair and is free of charge. There is no word limit on abstracts but please limit abstract length to one page, including title, authors, equations, and up to one expository figure. All abstracts will be considered for contributed talks; there will be no posters for SINM 2017.
Abstract submission is now closed.
This workshop will address the intersection of two trends in network science. On the one hand, real-world networks are increasingly annotated with rich metadata, including vertex or edge attributes or temporal information. Making sense of such data requires moving beyond simple models of network structure. On the other, hypotheses about network structure and the processes that create those patterns are increasingly sophisticated. The tools of statistical inference for network models offer a principled and effective approach for both understanding richly annotated network data and testing interesting network hypotheses.
In particular, probabilistic models are a quantitative approach that allow researchers both to infer complicated hidden structural patterns in existing data and generate synthetic data sets whose structure is statistically similar to real data. These models facilitate handling many of the challenges of understanding real data, including accounting for noise and missing values, and they connect theory with data by providing interpretable results. Statistical inference is thus a powerful and useful tool for modeling and understanding networks.
The development of new tools and their application to understand real systems is now a major community effort in network science. Despite their power and utility, however, these techniques are not as easy or approachable as simpler tools, like degree distributions, centrality scores, and clustering coefficients. Increasingly, new applications and richer data sets offer new opportunities for developing and applying the principled techniques of statistical inference to networks.
This satellite symposium will build on three successful previous satellites at NetSci2014, NetSci2015, and NetSci2016 by uniting theoretical and applied researchers, bringing together approaches from across network science, including machine learning, statistics, and physics. This broad cross-section of disciplines shares problems and even approaches, but each discipline brings a different perspective, emphasis and vocabulary. The purpose of this symposium is to provide a platform for cross-pollination of ideas and promote interdisciplinary interactions which leverage the diversity of approaches to a common set of problems.
There will be no published proceedings for this satellite.
Invited Speakers - Abstracts and Titles
Elizabeth Ogburn, Johns Hopkins
Karl Rohe, Wisconsin
Jennifer Neville, Purdue
Tamara Broderick, MIT
Bailey Fosdick, Colorado State
Dan Larremore, Santa Fe Institute
Aaron Clauset, Colorado