The minimax optimal excess-risk rate for pure ε-DP heavy-tailed SCO is characterized up to logarithmic factors, with a polynomial-time algorithm based on Lipschitz extensions of the empirical loss and a nearly matching lower bound.
High-probability minimax lower bounds
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
verdicts
UNVERDICTED 6representative citing papers
Realisable epsilon-contamination models for MNAR data yield minimax mean estimation rates that decompose into MCAR plus robust terms and remain consistent for Gaussian bases even as missingness and epsilon both tend to 1.
OGPO enables sample-efficient full-finetuning of generative control policies via off-policy critics and modified PPO, achieving SOTA on robot manipulation tasks while rescuing poorly initialized behavior cloning policies without expert data.
The authors instantiate a generalized-Fano framework using squared Hellinger distance to derive explicit Bayesian CVaR lower bounds for interactive decision problems including Gaussian bandits.
ULS provides minimax-optimal estimation of remaining-data parameters in machine unlearning with limited access and decomposes error into oracle plus unlearning cost terms.
Derives instance-specific lower bounds on sample complexity for rank-adaptive matrix estimation and proposes a least-squares plus universal singular-value-thresholding algorithm whose finite-sample error nearly matches those bounds.
citing papers explorer
-
Estimation beyond Missing (Completely) at Random
Realisable epsilon-contamination models for MNAR data yield minimax mean estimation rates that decompose into MCAR plus robust terms and remain consistent for Gaussian bases even as missingness and epsilon both tend to 1.