IIHE invited seminar: (Machine) Learning to Pivot with Adversarial Networks

Abstract: Many inference problems involve data generation processes that are not uniquely specified or are uncertain in some way. In a scientific context, the presence of several plausible data generation processes is often associated to the presence of systematic uncertainties. Robust inference is possible if it is based on a pivot -- a quantity whose distribution is invariant to the unknown value of the (categorical or continuous) nuisance parameters that parametrizes this family of generation processes. In this work, we introduce a flexible training procedure based on adversarial networks for enforcing the pivotal property on a predictive model. We derive theoretical results showing that the proposed algorithm tends towards a minimax solution corresponding to a predictive model that is both optimal and independent of the nuisance parameters (if that models exists) or for which one can tune the trade-off between power and robustness. Finally, we demonstrate the effectiveness of this approach with a toy example and an example from particle physics. 

 

Relevant links:

https://arxiv.org/abs/1611.01046

https://arxiv.org/abs/1703.03507 (for a recent and independent application to jet physics)

https://github.com/glouppe/notebooks/blob/master/Fair%20classifier%20with%20adversarial%20networks.ipynb