8 papers
Near-Optimal Stochastic Linear Bandits with Delay
Ofir Schlisselberg, Mengxiao Zhang, Yishay Mansour
We study stochastic linear bandits with delayed feedback under several delay models and establish near-optimal regret guarantees. Our results identify when delayed linear bandits e…
Mirror Descent Beyond Euclidean Stability: An Exponential Separation in Initialization Sensitivity
Shira Vansover-Hager, Matan Schliserman, Ofir Schlisselberg +1
Mirror Descent (MD) extends Gradient Descent (GD) beyond Euclidean geometry and has recently reappeared as a lens for KL-regularized policy optimization in reinforcement learning a…
Collaborating in Multi-Armed Bandits with Strategic Agents
Idan Barnea, Ofir Schlisselberg, Yishay Mansour
We study collaborative learning in multi-agent Bayesian bandit problems, where strategic agents collectively solve the same bandit instance. While multiple agents can accelerate le…
Online Learning in MDPs with Partially Adversarial Transitions and Losses
Ofir Schlisselberg, Tal Lancewicki, Yishay Mansour
We study reinforcement learning in MDPs whose transition function is stochastic at most steps but may behave adversarially at a fixed subset of steps per episode. This model ca…
The Hidden Cost of Approximation in Online Mirror Descent
Ofir Schlisselberg, Uri Sherman, Tomer Koren +1
Online mirror descent (OMD) is a fundamental algorithmic paradigm that underlies many algorithms in optimization, machine learning and sequential decision-making. The OMD iterates…
Improved Best-of-Both-Worlds Regret for Bandits with Delayed Feedback
Ofir Schlisselberg, Tal Lancewicki, Peter Auer +1
We study the multi-armed bandit problem with adversarially chosen delays in the Best-of-Both-Worlds (BoBW) framework, which aims to achieve near-optimal performance in both stochas…
The impact of allocation strategies in subset learning on the expressive power of neural networks
Ofir Schlisselberg, Ran Darshan
In traditional machine learning, models are defined by a set of parameters, which are optimized to perform specific tasks. In neural networks, these parameters correspond to the sy…
Delay as Payoff in MAB
Ofir Schlisselberg, Ido Cohen, Tal Lancewicki +1
In this paper, we investigate a variant of the classical stochastic Multi-armed Bandit (MAB) problem, where the payoff received by an agent (either cost or reward) is both delayed,…