Statistical Learning
Fast Optimal Locally Private Mean Estimation via Random Projections
Existing algorithms for this problem either incur sub-optimal error or have high communication and/or run-time complexity.
Large-Scale Distributed Learning via Private On-Device Locality-Sensitive Hashing
These authors contributed equally to this work.
Future-Dependent Value-Based Off-Policy Evaluation in POMDPs
Existing methods such as sequential importance sampling estimators suffer from the curse of horizon in POMDPs.
Should I Stop or Should I Go: Early Stopping with Heterogeneous Populations Hammaad Adam Massachusetts Institute of Technology Fan Yin
Existing methods that determine when to stop an experiment early are typically applied to the data in aggregate and do not account for treatment effect heterogeneity.
Overleaf Example
Experiments show that the proposed ReBalanced Adversarial Training (ReBA T) can attain good robustness and does not suffer from robust overfitting even after very long training.
