A Brief Survey of Deep Reinforcement Learning

Author(s): Kai Arulkumaran, Marc Peter Deisenroth, Miles Brundage, Anil Anthony Bharath

Abstract: Abstract—Deep reinforcement learning is poised to revolu-tionise the field of AI and represents a step towards buildingautonomous systems with a higher level understanding of thevisual world. Currently, deep learning is enabling reinforcementlearning to scale to problems that were previously intractable,such as learning to play video games directly from pixels. Deepreinforcement learning algorithms are also applied to robotics,allowing control policies for robots to be learned directly fromcamera inputs in the real world. In this survey, we begin withan introduction to the general field of reinforcement learning,then progress to the main streams of value-based and policy-based methods. Our survey will cover central algorithms indeep reinforcement learning, including the deepQ-network,trust region policy optimisation, and asynchronous advantageactor-critic. In parallel, we highlight the unique advantages ofdeep neural networks, focusing on visual understanding viareinforcement learning. To conclude, we describe several currentareas of research within the field.

Read the full paper here.