Back to All Events

Nathan Kallus on Seminar Series: Causal Inference and Data Science for Improved Policy

On May 29, 2024, DSI hosted the presentation titled “Debiased Inference on Functionals of Inverse Problems and Applications to Long-Term Causal Inference,” marking the third talk of our seminar series that encourages research collaboration among data scientists, causal inference experts, and applied researchers. This series is funded through the DSI Emergent Data Sciences Program competition, and co-led by Professors Linbo Wang (Departments of Statistical Sciences and Computer Science), Gustavo J. BobonisRaji Jayaraman, and Ismael Mourifié (Department of Economics) at the University of Toronto.


Nathan Kallus, Associate Professor, Cornell Tech, Cornell University

Featured Research

Prof. Kallus discusses how to make reliable inferences on causal effects in the presence of endogeneity using instruments and negative controls. Kallus and co-authors examine debiased estimators that are robust and asymptotically normal, enabled by a new adversarial learning method. Additionally, they address the issue of weak instruments in A/B tests on digital platforms, ensuring consistency with a jackknifed loss function.

Related Past Events