Research overview
Research by Topic
Explore six complementary areas spanning statistical foundations, learning, computation, and scientific applications. Papers are organized by their primary contribution and may appear in one additional area when the secondary contribution is equally substantive.
Reliable causal inference
Identification and causal-effect inference under invalid instruments, endogeneity, and hidden confounding.
Explore topic Learning across sourcesMulti-source learning
Transfer, federation, and stable learning across heterogeneous populations and environments.
Explore topic Statistical foundationsNonstandard inference
Valid confidence statements after selection, slow nuisance estimation, and other nonregular behavior.
Explore topic Statistical foundationsHigh-dimensional uncertainty quantification
Confidence intervals and testing when parameters rival or exceed observations.
Explore topic Scientific applicationsGenetics and health data
Methods driven by genomics, proteomics, biomarkers, and multi-center clinical data.
Explore topic Optimization and computationDistributionally Robust Learning and Optimization
Statistical formulations and algorithms for distributional robustness and structured nonconvex problems.
Explore topic