wxmcp3-CSDN博客

原创【读博客/翻译】A Visual Guide to Evolution Strategies 进化策略的视觉指南

进化策略的视觉指南 A Visual Guide to Evolution Strategiesbrief博客地址链接这个图首先就好有趣！Survival of the fittest.物竞天择，适者生存下面开始正文：在这篇文章中，我借助一些直观的例子来解释进化策略（ES）是如何工作的。我尽量让方程式保持轻松，如果读者希望了解更多细节，我提供了原始文章的链接。这是一系列文章的第一篇，我计划在这里展示如何将这些算法应用到从MNIST、OpenAI Gym、Roboschool到PyBullet

2020-11-12 10:03:18 810 1

原创【文献阅读】PS 综述 in Continuous Action Domains: an Overview

Policy Search in Continuous Action Domains: an OverviewAbstractContinuous action policy search is currently the focus of intensive research, driven both by the recent success of deep reinforcement learning algorithms and the emergence of competitors base

2020-10-03 09:55:43 958 1

原创【文献阅读】ES as a Scalable Alternative to RL（OpenAI 17）

Evolution Strategies as a Scalable Alternative to Reinforcement Learningbrief文章链接该文章是 Open AI 17年发布的，目前有300+的引用量。Abstract【开篇明意】We explore the use of Evolution Strategies (ES), a class of black box optimization algorithms, as an alternative to popular

2020-09-23 20:24:50 1670

原创【文献阅读】17年进化算法和DRL结合的文章

Improving Exploration in Evolution Strategies for Deep Reinforcement Learning via a Population of Novelty-Seeking AgentsBrief目前该文已经有了上百的引用量，还是有点厉害。文章地址链接代码链接 code作者来自佛罗里达大学和openAIAbstract【开篇明义】Evolution strategies (ES) are a family of black-box opt

2020-09-14 21:20:11 1219

原创【文献阅读】Soft Actor-Critic（ERL的前一版）

Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic ActorBrief文章链接paper代码链接code作者Sergey Levine来自加州伯克利分校UC BerkeleyAbstract

2020-09-04 16:51:38 687

原创【文献阅读】RL经典：Benchmarking Deep Reinforcement Learning for Continuous Control

Benchmarking Deep Reinforcement Learning for Continuous ControlBriefpaper 链接paper开源代码code引用量 300++作者Pieter Abbeel来自UC BerkeleyAbstract鉴于RL领域具有很强的时效性，这篇2016年开源的论文在当时应该是非常轰动的。不过时至今日，...

2020-08-31 10:02:01 1155

原创【文章阅读】BN(2015)理解Batch Normalization批标准化

Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate ShiftBrief该论文是关于 BN，引用量300+论文链接paperPieter Abbeel Sergey Levine 作者依然是两位大佬

2020-07-19 21:23:40 836

原创【文献阅读】进化导向的策略梯度RL

Evolution-Guided Policy Gradient in Reinforcement LearningBrief文章链接 paper代码链接code作者 Kagan Tumer作者另外一个作者是因特尔人工智能研究院的。Intel AI · Ai-Lab俄勒冈州立大学协作机器人与智能系统研究所Collaborative Robotics and Intelligent Systems Institute Oregon State UniversityAbstract文章是在.

2020-07-15 10:54:19 1155

原创【文章阅读】连续空间增量式RL(策略松弛和加权权重)

【文章阅读】Incremental Reinforcement Learning in Continuous Spaces via Policy Relaxation and Importance WeightingBriefAbstractIntroduction第一段：第二段：第三段：第四段：第五段：第六段：第七段：BackgroundA. Reinforcement Learning in Continuous Spaces连续空间强化学习B. Related Work相关工作第三章方法设计INCRE

2020-07-14 20:33:08 681

wxmcp3的博客