by Satwik Kansal | Released August 2018 | ISBN: 9781788995368

Train your agent using Reinforcement Learning with Tensorflow's neural networks, OpenAI Gym and Python, to make it smarterAbout This VideoPractical training in the Reinforcement Learning architecture for training agentsWork with important open source Reinforcement Learning frameworks to get an in-depth knowledge of its functionsA Production-ready approach to training Reinforcement Learning agents in Tensorflow to take on real-world projectsIn DetailYou've probably heard of Deepmind's AI playing games and getting really good at playing them (like AlphaGo beating the Go world champion). Such agents are built with the help of a paradigm of machine learning called “Reinforcement Learning” (RL). In this course, you'll walk through different approaches to RL. You'll move from a simple Q-learning to a more complex, deep RL architecture and implement your algorithms using Tensorflow's Python API. You'll be training your agents on two different games in a number of complex scenarios to make them more intelligent and perceptive.By the end of this course, you'll be able to implement RL-based solutions in your projects from scratch using Tensorflow and Python.The code bundle for this video course is available at: https://github.com/PacktPublishing/-Hands-on-Reinforcement-Learning-with-TensorFlow Show and hide more Publisher Resources Download Example Code

1. Chapter 1 : Understanding the Reinforcement Learning Framework
   • The Course Overview 00:04:35
   • Introduction to Reinforcement Learning 00:05:39
   • Common RL Tasks and the Reinforcement Process 00:07:19
   • Setting Up Environments Using Open AI’s Gym Framework 00:05:57
2. Chapter 2 : Training a Smartcab Using Q-Learning
   • The Taxi-v2 Environment 00:10:25
   • Operating Taxi-v2 Using a Dumb Agent 00:05:44
   • Introducing Reinforcement Q-Learning 00:06:24
   • Implementing Q-Learning 00:07:46
   • Q-Learning Agent in Action 00:09:33
3. Chapter 3 : Balancing a Cartpole Using Q-Networks
   • The Cartpole Environment 00:08:12
   • Introducing Q-Networks 00:07:06
   • TensorFlow Basics 00:11:05
   • Implementing Q-Network 00:09:19
   • Q-Network Agent in Action 00:04:21
4. Chapter 4 : Deep Reinforcement Learning with TensorFlow
   • Introducing Deep Q-Networks 00:06:49
   • The DQN Training Algorithm 00:04:49
   • Implementing DQN 00:12:08
   • DQN in Action 00:17:45
   • Dueling Double DQN 00:12:59
5. Chapter 5 : Getting Production-Ready with TensorFlow
   • Logging, Saving, and Visualizing 00:10:04
   • Structuring the Code Base 00:19:06
   • Debugging and Some Nice Practices in TensorFlow 00:17:21
   • TensorFlow on Multiple Devices 00:13:49
   • Next Steps 00:03:51
Show and hide more