The sample-averaged gradient of the CSO objective is biased due to its nested structure, and therefore requires a high sample complexity for convergence. We introduce a general stochastic extrapolation technique that effectively reduces the bias.

We demonstrate that this assumption imposes performance limitations in particular on difficult problem instances.

The hope is to incur privacy loss only on positive responses.

We propose the T arget Charging T echnique (TCT), a unified privacy analysis framework for interactive settings where a sensitive dataset is accessed multiple times using differentially private algorithms. Unlike traditional composition, where privacy guarantees deteriorate quickly with the number of accesses, TCT allows computations that don't hit a specified target, often the vast majority, to be essentially free (while incurring instead a small overhead on those that do hit their targets). TCT generalizes tools such as the sparse vector technique and top-k selection from private candidates and extends their remarkable privacy enhancement benefits from noisy Lipschitz functions to general private algorithms.

Algorithmic stability is a major theme in learning theory, where seminal results have firmly established its close relationship with generalization. Recent research has further highlighted the intricate interplay between stability and additional properties of interest beyond statistical generalization.

In this work, we study the low-rank MDPs with adversarially changed losses in the full-information feedback setting.