Reinforcement Learning
Meta-Gradient Reinforcement Learning with an Objective Discovered Online
Each algorithm optimises its parameters with respect to an objective, such as Q-learning or policy gradient, that defines its semantics.
Sample-Efficient Reinforcement Learning Is Feasible for Linearly Realizable MDPs with Limited Revisiting
This type of MDPs is commonly referred to as linear MDPs.
Almost Optimal Model-Free Reinforcement Learning via Reference-Advantage Decomposition
Our regret bound improves upon the results of [Jin et al., 2018] and
Reinforcement Learning in Linear MDPs: Constant Regret and Representation Selection
We study the role of the representation of state-action value functions in regret minimization in finite-horizon Markov Decision Processes (MDPs) with linear structure.
Variance-Reduced Off-Policy TDC Learning: Non-Asymptotic Convergence Analysis
Recently, several work proposed to apply the variance reduction technique developed in the stochastic optimization literature to reduce the variance of TD learning.