Reinforcement learning, mathematically described by Markov Decision Problems, may be
approached either through dynamic programming or policy search. Actor-critic algorithms …

Reinforcement learning consists of finding policies that maximize an expected cumulative
long-term reward in a Markov decision process with unknown transition probabilities and …

We focus on policy search in reinforcement learning problems over continuous spaces,
where the value is defined by infinite-horizon discounted reward accumulation. This is the …

We consider policy evaluation in infinite-horizon discounted Markov decision problems with
continuous compact state and action spaces. We reformulate this task as a compositional …

This paper considers policy search in continuous state-action reinforcement learning
problems. Typically, one computes search directions using a classic expression for the …

Many machine learning tasks can be formulated as a stochastic compositional optimization
(SCO) problem such as reinforcement learning, AUC maximization, and meta-learning …

Reinforcement learning is a framework to optimize an agent's policy using rewards that are
revealed by the system as a response to an action. In its standard form, reinforcement …

Estimating a policy that maps states to actions is a central problem in reinforcement learning.
Traditionally, policies are inferred from the so called value functions (VFs), but exact VF …

Recently, more and more researchers have used deep reinforcement learning (DRL) to
solve the UAV local path planning problem. However, existing DRL didn't consider the …

In this work, we address optimization problems where the objective function is a nonlinear
function of an expected value, ie, compositional stochastic strongly convex programs. We …