arXiv research

Cyclical learning rate improves neural machine translation performance.

Policy gradient algorithm with variable learning rates achieves near-optimal performance in multi-arm bandit problems.

LALR adapts learning rate for faster convergence in regression and neural nets.

Learning Rate (LR) is an important hyper-parameter to tune for effective training of deep neural networks (DNNs). Even for the baseline of a constant learning rate, it is non-trivial to choose a good constant value for training a DNN. Dynamic learning rates involve multi-step tuning of LR values at various stages of th…

New guarantees for ERM with adaptively collected data.

Study proposes a new approval policy for ML-based medical devices to prevent gradual performance degradation.

Minimal assumptions analysis of Q-learning with time-varying policies.

New methods estimate policy value and gradients for deterministic policies from off-policy data.

Study on policy gradient for stochastic bandits using diffusion approximation.

Robotic table tennis learns efficient policies to return balls at 100Hz.

Develops DPG methods for continuous-time RL with deterministic policies.

Entropy regularized algorithms such as Soft Q-learning and Soft Actor-Critic, recently showed state-of-the-art performance on a number of challenging reinforcement learning (RL) tasks. The regularized formulation modifies the standard RL objective and thus generally converges to a policy different from the optimal gree…

Counterfactual policy evaluation improves autonomous driving policies' generalization.

A drone catches another agile drone using competitive reinforcement learning.

New reinforcement learning algorithm achieves instance-optimal sample complexity.

Linear Q-learning converges to a bounded set without divergence.

This paper tackles robust policy learning under concept drifts, improving upon existing methods.

Trust region policy optimization (TRPO) is a popular and empirically successful policy search algorithm in Reinforcement Learning (RL) in which a surrogate problem, that restricts consecutive policies to be 'close' to one another, is iteratively solved. Nevertheless, TRPO has been considered a heuristic algorithm inspi…

Single-timescale actor-critic finds globally optimal policy.

The paper tackles adaptive policy selection to maximize social welfare, achieving optimal regret bounds.

Tabular Q-learning outperforms advanced RL methods in monetary policy.

Paper achieves $ε^{-2}$ sample complexity for actor-critic methods with minimal assumptions.

New method estimates and optimizes policy differences using orthogonal learning.

Softmax policy gradient methods converge at $O(1/t)$ rate with constants depending on problem and initialization.

Unified framework for stable RL learning with theoretical guarantees.

The paper tackles robust policy learning in MDPs using statistical methods.

Adaptive OMD reduces variance in learning optimal strategies for imperfect information games.

Unified framework for risk-aware policy learning in contextual bandits.

Rate change calculations in the literature involve deterministic methods that measure the change in premium for a given policy. The definition of rate change as a statistical parameter is proposed to address the stochastic nature of the premium charged for a policy. It promotes the idea that rate change is a property o…

Algorithm learns to solve tasks by gradually expanding policy space.

Large deviations theory applied to policy gradient methods.

The paper analyzes reinforcement learning methods for estimating weights and quality functions with fast convergence rates.

Sample efficiency is critical in solving real-world reinforcement learning problems, where agent-environment interactions can be costly. Imitation learning from expert advice has proved to be an effective strategy for reducing the number of interactions required to train a policy. Online imitation learning, which inter…

The paper tackles robust policy learning from multiple data sources.

Parrot learns optimal cache replacement policies using imitation learning.

Study introduces a new framework for policy learning without positivity assumption.

This paper investigates to what extent one can improve reinforcement learning algorithms. Our study is split in three parts. First, our analysis shows that the classical asymptotic convergence rate $O(1/\sqrt{N})$ is pessimistic and can be replaced by $O((\log(N)/N)^β)$ with $\frac{1}{2}\leq β\leq 1$ and $N$ the number…

The paper develops a method to learn cost-optimal sequential testing policies from retrospective data.

Study minimax rates for online learning with time-varying dynamics.

Proposes DR algorithms for distributionally robust off-policy evaluation and learning.

We decode latent states in Block MDPs and learn near-optimal policies.

Study uses ML to analyze how interest rates affect fund returns, finding gradient boosting is effective.

Develops a new reinforcement learning framework for complex control problems.

New method for RL with general utilities using variational policy gradient.

Paper tackles efficient policy gradient estimation from off-policy data.

The paper proposes an interpretable off-policy learning algorithm for medical treatments.

Policy gradient methods achieve linear convergence in simple MDPs.

Study Federated RL with diverse constraints, proposing new optimization methods.