Most policy search (PS) algorithms require thousands of training episodes to find an
effective policy, which is often infeasible with a physical robot. This survey article focuses on …

Offline reinforcement learning (RL) refers to the problem of learning policies entirely from a
batch of previously collected data. This problem setting is compelling, because it offers the …

Recent work has shown that supervised learning alone, without temporal difference (TD)
learning, can be remarkably effective for offline RL. When does this hold true, and which …

Designing effective model-based reinforcement learning algorithms is difficult because the
ease of data generation must be weighed against the bias of model-generated data. In this …

Sequential decision making, commonly formalized as Markov Decision Process (MDP)
optimization, is an important challenge in artificial intelligence. Two key approaches to this …

A generative recurrent neural network is quickly trained in an unsupervised manner to
model popular reinforcement learning environments through compressed spatio-temporal …

Abstract Model-based reinforcement learning (RL) algorithms can attain excellent sample
efficiency, but often lag behind the best model-free algorithms in terms of asymptotic …

PowerPoint 演示文稿 Page 1 Uncertainty in Deep Learning Yarin Gal 2018.7.29 Page 2 Page
3 Different Uncertainties Two main types of uncertainty, often confused by practitioners, but …

Model-free reinforcement learning (RL) methods are succeeding in a growing number of
tasks, aided by recent advances in deep learning. However, they tend to suffer from high …

There is growing interest in combining model-free and model-based approaches in
reinforcement learning with the goal of achieving the high performance of model-free …