We study the problem of multi-armed bandits with ε-global Differential Privacy (DP). First, we
prove the minimax and problem-dependent regret lower bounds for stochastic and linear …

Motivated by the wide adoption of reinforcement learning (RL) in real-world personalized
services, where users' sensitive and private information needs to be protected, we study …

The offline reinforcement learning (RL) problem is often motivated by the need to learn data-
driven decision policies in financial, legal and healthcare applications. However, the learned …

We address differentially private stochastic bandits. We present two (near)-optimal
Thompson Sampling-based learning algorithms: DP-TS and Lazy-DP-TS. The core idea in …

We study regret minimization in finite horizon tabular Markov decision processes (MDPs)
under the constraints of differential privacy (DP). This is motivated by the widespread …

Bandits serve as the theoretical foundation of sequential learning and an algorithmic
foundation of modern recommender systems. However, recommender systems often rely on …

A matroid bandit is the online version of combinatorial optimization on a matroid, in which
the learner chooses $ K $ actions from a set of $ L $ actions that can form a matroid basis …

We consider the problem of online classification under a privacy constraint. In this setting a
learner observes sequentially a stream of labelled examples $(x_t, y_t) $, for $1\leq t\leq T …

In this work we first show that the classical Thompson sampling algorithm for multi-arm
bandits is differentially private as-is, without any modification. We provide per-round privacy …

This thesis focuses on two classes of learning problems in stochastic multi-armed bandits
(MAB): graphical bandits and private bandits. Different from the basic MAB setting where the …