Causal Inference and Prediction for Network Data (24w5244)


Tianxi Li (University of Minnesota)

Eric D. Kolaczyk (McGill University)

(University of Michigan)

Elizabeth Ogburn (Johns Hopkins university)


The Banff International Research Station will host the “Causal Inference and Prediction for Network Data” workshop in Banff from August 18 - 23, 2024.

Advances in data collection have made network data available in many areas, including social sciences, biological sciences, and engineering. Understanding and modeling network structure and conducting rigorous statistical inference to assess uncertainty can provide crucial insights into the dynamics and interaction mechanisms of the system. This workshop, in particular, focuses on modeling and inferring causal relations in network data and leveraging these model inference for predictions.

Causal inference between variables based on observations in a network has been an extremely challenging problem, requiring adapting existing inference frameworks to networks. The workshop will start by covering recent advances in this setting. Yet another challenging problem emerging from network data is the inference of causal relations between attributes and the network formulation. Such inference requires inventing new models for the corresponding data-generating processes and new tools to analyze these statistical models. The workshop would also cover advances in this setting. An even more challenging setting would be transitioning from a single network or a single pair of variables to more complicated data structures, such as the setting of multiple and dependent networks, the interplay between network formation, and high-dimensional multivariate data. Moreover, in these complex data sets, leveraging causality for predictions is a practically valuable problem. These problems stem from a wide range of motivating applications, especially in neuroscience, biology, and the social sciences. The workshop will bring researchers from theory, computation, and different applications together, to help theoreticians and methodologists focus on real problems, and to alert application researchers to the newest developments in methods.

The Banff International Research Station for Mathematical Innovation and Discovery (BIRS) is a collaborative Canada-US-Mexico venture that provides an environment for creative interaction as well as the exchange of ideas, knowledge, and methods within the Mathematical Sciences, with related disciplines and with industry. The research station is located at The Banff Centre in Alberta and is supported by Canada’s Natural Science and Engineering Research Council (NSERC), the U.S. National Science Foundation (NSF), Alberta’s Advanced Education and Technology, and Mexico’s Consejo Nacional de Ciencia y Tecnología (CONACYT).