Incentivizing exploration in reinforcement learning with deep predictive models

Stadie, Bradly C., Sergey Levine, and Pieter Abbeel. "Incentivizing exploration in reinforcement learning with deep predictive models." arXiv preprint arXiv:1507.00814 (2015).

作者通过模拟(状态，动作)的不确定性，从而修改reward，帮助agent进行探索。作者说用了他们的方法不用进行随机探索。该方法比较通用，适用于多种RL模型，但是要训练auto-encoder，所以也稍微有点繁琐。

实用指数：3颗星

理论指数：1颗星

创新指数：4颗星

时间： 2024-08-07 12:30:54

Incentivizing exploration in reinforcement learning with deep predictive models的相关文章

(zhuan) Deep Reinforcement Learning Papers

Deep Reinforcement Learning Papers A list of recent papers regarding deep reinforcement learning. The papers are organized based on manually-defined bookmarks. They are sorted by time to see the recent papers first. Any suggestions and pull requests

论文笔记之：Deep Reinforcement Learning with Double Q-learning

Deep Reinforcement Learning with Double Q-learning Google DeepMind Abstract 主流的 Q-learning 算法过高的估计在特定条件下的动作值.实际上,之前是不知道是否这样的过高估计是 common的,是否对性能有害,以及是否能从主体上进行组织.本文就回答了上述的问题,特别的,本文指出最近的 DQN 算法,的确存在在玩 Atari 2600 时会 suffer from substantial overestimation

(转) Deep Learning in a Nutshell: Reinforcement Learning

Deep Learning in a Nutshell: Reinforcement Learning Share: Posted on September 8, 2016by Tim Dettmers No CommentsTagged Deep Learning, Deep Neural Networks, Machine Learning,Reinforcement Learning This post is Part 4 of the Deep Learning in a Nutshel

(转) Deep Reinforcement Learning: Playing a Racing Game

Byte Tank Posts Archive Deep Reinforcement Learning: Playing a Racing Game OCT 6TH, 2016 Agent playing Out Run, session 201609171218_175epsNo time limit, no traffic, 2X time lapse Above is the built deep Q-network (DQN) agent playing Out Run, trained

Deep Reinforcement Learning 基础知识（DQN方面）

Introduction 深度增强学习Deep Reinforcement Learning是将深度学习与增强学习结合起来从而实现从Perception感知到Action动作的端对端学习的一种全新的算法.简单的说,就是和人类一样,输入感知信息比如视觉,然后通过深度神经网络,直接输出动作,中间没有hand-crafted工作.深度增强学习具备使机器人实现完全自主的学习一种甚至多种技能的潜力. 虽然将深度学习和增强学习结合的想法在几年前就有人尝试,但真正成功的开端是DeepMind在NIPS 201

repost: Deep Reinforcement Learning

From: http://wanghaitao8118.blog.163.com/blog/static/13986977220153811210319/ accessed 2016-03-10 深度强化学习(Deep Reinforcement Learning)的资源 Google的Deep Mind团队2013年在NIPS上发表了一篇牛x闪闪的文章,亮瞎了好多人眼睛,不幸的是我也在其中.前一段时间收集了好多关于这方面的资料,一直躺在收藏夹中,目前正在做一些相关的工作(希望有小伙伴一起交流)

(转) Playing FPS games with deep reinforcement learning

Playing FPS games with deep reinforcement learning 博文转自:https://blog.acolyer.org/2016/11/23/playing-fps-games-with-deep-reinforcement-learning/ When I wrote up 'Asynchronous methods for deep learning' last month, I made a throwaway remark that after

论文笔记之：Dueling Network Architectures for Deep Reinforcement Learning

Dueling Network Architectures for Deep Reinforcement Learning ICML 2016 Best Paper Google DeepMind Abstract: 本文是 ICML 2016 的最佳论文之一,又是出自 Google DeepMind. 最近几年,在 reinforcement learning 上关于 deep representation 有取得了很大的成功.然而,许多这些应用都是利用传统的网络架构,例如:神经网络,LSTM

Playing Atari with Deep Reinforcement Learning

这是一篇论文,原地址在: https://arxiv.org/abs/1312.5602 我属于边看便翻译,边理解,将他们记录在这里: Abstract: 我们提出了第一个深学习模型,成功地学习控制策略直接从高维感官输入使用强化学习.该模型是一个卷积神经网络,用Q-学习的变体训练,其输入是原始像素,其输出是估计未来的值函数.我们运用我们的方法在Atari 2600 游戏中测试,没有调整结构或学习的算法.我们发现它比所有之前的方法都好,比人类专家玩得都厉害. 1 Introduction 直接从高