|
所在平台: Udemy |
课程主页: https://www.udemy.com/course/fast-deep-reinforcement-learning/
课程评论：没有评论
**课程名称：** 快速学习深度强化学习 **课程概述：** 本课程旨在教授学员如何使用行业领先的深度强化学习框架 Ray RLlib，以最快的方式（数小时而非数天）构建和原型化深度强化学习（Deep RL）代理。课程将从基础概念讲起，包括环境、动作、累积奖励最大化等强化学习核心概念，并涵盖工业应用案例。 学员将学习如何控制 OpenAI Gym 环境中的代理，并利用 Ray RLlib 解决这些环境中的问题。课程还重点介绍如何使用 Tensorboard 可视化代理的学习行为，以及如何保存和加载训练好的代理。此外，学员还将学习如何为特定问题选择最佳的深度强化学习算法。 课程采用“边做边学”的学习方法，通过实际代码编写和引导式编码练习，帮助学员巩固所学知识。整个课程时长约为 4-8 小时，包括简短的视频（平均 6 分钟）和实践练习。 **课程亮点：** * **高效学习：** 使用 Ray RLlib 框架，大幅缩短深度强化学习代理的原型开发时间。 * **从零开始：** 即使是初学者，也能轻松掌握强化学习核心概念和 Ray RLlib 的使用。 * **实践导向：** 通过编码练习和实际项目（训练机器人行走），掌握可迁移的实用技能。 * **可视化工具：** 利用 Tensorboard 直观了解代理的学习过程，便于排查问题。 * **专家推荐：** 课程内容被专家评价为“将复杂的 RL 概念分解成易于理解的小部分”以及“对 RL 概念及其在 RLlib 中的映射进行了出色的介绍”。 * **短视频内容：** 视频长度适中，内容精炼，学习效率高。 * **高质量字幕：** 提供高质量的英语字幕，支持多语言学习者。 **核心学习内容：** * 强化学习核心概念（环境、动作、累积奖励最大化等） * 强化学习在工业中的应用案例 * 判断何时使用强化学习而非传统方法 * 控制 OpenAI Gym 环境中的代理 * 使用 Ray RLlib 解决 OpenAI Gym 问题 * 使用 Tensorboard 可视化学习行为 * 保存和加载训练好的代理 * 选择合适的深度强化学习算法 **学习方式：** * 视频讲授 * 编码实践 * 课堂测验 * 引导式编码练习（围绕训练机器人行走项目）
Used Keras or PyTorch? These frameworks make it easy to build Deep Neural Networks.New Deep Reinforcement Learning frameworks like Ray RLlib make it similarly easy to build Deep RL agents. Using Ray RLlib, it's possible to prototype Deep RL agents in hours instead of days. This course will show you how to do that. We will start from scratch, and after a few evenings of lessons and exercises, you will be able to code powerful Deep RL agents using Ray RLlib to solve various OpenAI Gym environments. This is the fastest way to get a feeling for Deep Reinforcement Learning. We will cover the following topics in the course.Core concepts of Reinforcement Learning like environment, action, cumulative reward maximization, etc.Case studies of Reinforcement Learning applications in the industryHow to decide whether to use Reinforcement Learning or conventional methods for a given learning taskHow to control agents inside OpenAI Gym environment (a Gym environment is just a simulation of a learning task)How to use the industry-leading Deep Reinforcement Learning framework Ray RLlib to solve OpenAI Gym environmentsVisualizing the agent's learning behavior in Tensorboard (useful for troubleshooting)Saving and using the trained agentHow to choose the best Deep RL algorithm for a given problemThe course follows the learning-by-doing approach. This means that I will write code to solve an example problem and explain the concepts along the way in the right context. In the guided coding exercises, you will be challenged to apply what you have learned. This will ensure that you are learning applicable skills.Here are some other features of this course.The course consists of short videos with no fluff (on average, 6 minutes long). The entire course can be completed in 4 to 8 hours (including exercises).The videos have high-quality English captions.The lessons are often followed by quizzes and coding exercises so that you can test your knowledge.The exercises are part of an overarching project, where we teach a robot how to walk. We will record a video of this agent at the end of the course, making it easy to share your new skills with others (if you wish).This course was reviewed by a few experts and this is what they said:"This course broke down complex RL concepts into small pieces that I could easily understand" - Martin Musiol, Managing Data Scientist at IBM"Brilliant introduction to RL concepts and how they map to RLlib." - Jules Damji, Developer Advocate at Anyscale (creators of Ray RLlib)
