Experience replay lets online reinforcement learning agents remember and
reuse experiences from the past. In prior work, experience transitions were
uniformly sampled from a replay memory. However, this approach simply replays
transitions at the same frequency that they were originally experienced,
regardless of their significance. In this paper we develop a framework for
prioritizing experience, so as to replay important transitions more frequently,
and therefore learn more efficiently. We use prioritized experience replay in
Deep Q-Networks (DQN), a reinforcement learning algorithm that achieved
human-level performance across many Atari games. DQN with prioritized
experience replay achieves a new state-of-the-art, outperforming DQN with
uniform replay on 41 out of 49 games.
%0 Generic
%1 schaul2015prioritized
%A Schaul, Tom
%A Quan, John
%A Antonoglou, Ioannis
%A Silver, David
%D 2015
%K final reinforcement_learning thema:double_dqn
%T Prioritized Experience Replay
%U http://arxiv.org/abs/1511.05952
%X Experience replay lets online reinforcement learning agents remember and
reuse experiences from the past. In prior work, experience transitions were
uniformly sampled from a replay memory. However, this approach simply replays
transitions at the same frequency that they were originally experienced,
regardless of their significance. In this paper we develop a framework for
prioritizing experience, so as to replay important transitions more frequently,
and therefore learn more efficiently. We use prioritized experience replay in
Deep Q-Networks (DQN), a reinforcement learning algorithm that achieved
human-level performance across many Atari games. DQN with prioritized
experience replay achieves a new state-of-the-art, outperforming DQN with
uniform replay on 41 out of 49 games.
@misc{schaul2015prioritized,
abstract = {Experience replay lets online reinforcement learning agents remember and
reuse experiences from the past. In prior work, experience transitions were
uniformly sampled from a replay memory. However, this approach simply replays
transitions at the same frequency that they were originally experienced,
regardless of their significance. In this paper we develop a framework for
prioritizing experience, so as to replay important transitions more frequently,
and therefore learn more efficiently. We use prioritized experience replay in
Deep Q-Networks (DQN), a reinforcement learning algorithm that achieved
human-level performance across many Atari games. DQN with prioritized
experience replay achieves a new state-of-the-art, outperforming DQN with
uniform replay on 41 out of 49 games.},
added-at = {2019-11-17T22:17:21.000+0100},
author = {Schaul, Tom and Quan, John and Antonoglou, Ioannis and Silver, David},
biburl = {https://www.bibsonomy.org/bibtex/2be47ebfc4c5eb8f06b374fe632a3e236/jan.hofmann1},
description = {[1511.05952] Prioritized Experience Replay},
interhash = {db6e4402b0807938aae61afc45b70f73},
intrahash = {be47ebfc4c5eb8f06b374fe632a3e236},
keywords = {final reinforcement_learning thema:double_dqn},
note = {cite arxiv:1511.05952Comment: Published at ICLR 2016},
timestamp = {2019-12-09T10:13:58.000+0100},
title = {Prioritized Experience Replay},
url = {http://arxiv.org/abs/1511.05952},
year = 2015
}