Playing Atari with Deep Reinforcement Learning

Published in

arXiv.org(2013)

External Links:

Generate Graph

TL;DR

This work presents the first deep learning model to successfully learn control policies directly from high-dimensional sensory input using reinforcement learning, which outperforms all previous approaches on six of the games and surpasses a human expert on three of them.

Abstract

We present the first deep learning model to successfully learn control policies directly from high-dimensional sensory input using reinforcement learning. The model is a convolutional neural network, trained with a variant of Q-learning, whose input is raw pixels and whose output is a value function estimating future rewards. We apply our method to seven Atari 2600 games from the Arcade Learning Environment, with no adjustment of the architecture or learning algorithm. We find that it outperforms all previous approaches on six of the games and surpasses a human expert on three of them.

Authors

K. Kavukcuoglu

26 papers

Daan Wierstra

12 papers

David Silver

21 papers

References30 items

A Neuroevolution Approach to General Atari Game Playing

Bayesian Learning of Recursively Factored Environments

New types of deep neural network learning for speech recognition and related applications: an overview

Speech recognition with deep recurrent neural networks

Sketch-Based Linear Value Function Approximation

Playing Atari with Deep Reinforcement Learning

Published in

arXiv.org(2013)

External Links:

Generate Graph

TL;DR

Abstract

Authors

K. Kavukcuoglu

26 papers

Daan Wierstra

12 papers

David Silver

21 papers

References30 items

A Neuroevolution Approach to General Atari Game Playing

Bayesian Learning of Recursively Factored Environments

New types of deep neural network learning for speech recognition and related applications: an overview

Speech recognition with deep recurrent neural networks

Sketch-Based Linear Value Function Approximation

Volodymyr Mnih

6 papers

Alex Graves

11 papers

Ioannis Antonoglou

8 papers

Martin A. Riedmiller

7 papers

ImageNet classification with deep convolutional neural networks

Pedestrian Detection with Unsupervised Multi-stage Feature Learning

Investigating Contingency Awareness Using Atari 2600 Games

The Arcade Learning Environment: An Evaluation Platform for General Agents

Context-Dependent Pre-Trained Deep Neural Networks for Large-Vocabulary Speech Recognition

Deep auto-encoder neural networks in reinforcement learning

Toward Off-Policy Learning Control with Function Approximation

Rectified Linear Units Improve Restricted Boltzmann Machines

Convergent Temporal-Difference Learning with Arbitrary Smooth Function Approximation

What is the best multi-stage architecture for object recognition?

Learning multiple layers of representation

Neural Fitted Q Iteration - First Experiences with a Data Efficient Neural Reinforcement Learning Method

Reinforcement Learning with Factored States and Actions

Reinforcement Learning: : An Introduction

Convolutional networks for images, speech, and time series

Why did TD-Gammon Work?

Residual Algorithms: Reinforcement Learning with Function Approximation

Temporal Difference Learning and TD-Gammon

Prioritized Sweeping: Reinforcement Learning with Less Data and Less Time

Machine Learning for Aerial Image Labeling

Actor-Critic Reinforcement Learning with Energy-Based Policies

Reinforcement Learning: An Introduction

An Analysis of Temporal-Difference Learning with Function Approximation

Reinforcement Learning for Robots Using Neural Networks

Field of Study

Computer Science

Journal Information

Name

ArXiv

Volume

abs/1312.5602

Venue Information

Name

arXiv.org

Type

URL

https://arxiv.org

Alternate Names

ArXiv