Welcome to Cutting-Edge AI!

This is technically*Deep Learning in Python part 11* of my deep learning series, and my *3rd* reinforcement learning course.

Deep Reinforcement Learning is actually the combination of 2 topics:**Reinforcement Learning** and **Deep Learning (Neural Networks)**.

While both of these have been around for quite some time, it’s only been recently that Deep Learning has really taken off, and along with it, Reinforcement Learning.

The maturation of deep learning has propelled advances in reinforcement learning, which has been around since the 1980s, although some aspects of it, such as the Bellman equation, have been for much longer.

Recently, these advances have allowed us to showcase just how powerful reinforcement learning can be.

We’ve seen how**AlphaZero** can master the game of Go using only self-play.

This is just a few years after the original AlphaGo already beat a world champion in Go.

We’ve seen real-world robots learn how to walk, and even recover after being kicked over, despite only being trained using simulation.

Simulation is nice because it doesn’t require actual hardware, which is expensive. If your agent falls down, no real damage is done.

We’ve seen real-world robots learn hand dexterity, which is no small feat.

Walking is one thing, but that involves coarse movements. Hand dexterity is complex - you have many degrees of freedom and many of the forces involved are extremely subtle.

Imagine using your foot to do something you usually do with your hand, and you immediately understand why this would be difficult.

Last but not least - video games.

Even just considering the past few months, we’ve seen some amazing developments. AIs are now beating professional players in**CS:GO** and **Dota 2**.

So what makes this course different from the first two?

Now that we know deep learning works with reinforcement learning, the question becomes: how do we improve these algorithms?

This course is going to show you a few different ways: including the powerful**A2C (Advantage Actor-Critic)** algorithm, the **DDPG (Deep Deterministic Policy Gradient)** algorithm, and **evolution strategies**.

Evolution strategies is a new and fresh take on reinforcement learning, that kind of throws away all the old theory in favor of a more "black box" approach, inspired by biological evolution.

What’s also great about this new course is the variety of environments we get to look at.

First, we’re going to look at the classic**Atari** environments. These are important because they show that reinforcement learning agents can learn based on images alone.

Second, we’re going to look at**MuJoCo**, which is a physics simulator. This is the first step to building a robot that can navigate the real-world and understand physics - we first have to show it can work with simulated physics.

Finally, we’re going to look at**Flappy Bird**, everyone’s favorite mobile game just a few years ago.

What do you get if you sign up for the VIP version of this course? A brand new*exclusive* section covering an entirely new algorithm: **TD3**! As usual, both theory and code for this powerful state-of-the-art algorithm are provided.

Thanks for reading, and I’ll see you in class!

Suggested Prerequisites:

This is technically

Deep Reinforcement Learning is actually the combination of 2 topics:

While both of these have been around for quite some time, it’s only been recently that Deep Learning has really taken off, and along with it, Reinforcement Learning.

The maturation of deep learning has propelled advances in reinforcement learning, which has been around since the 1980s, although some aspects of it, such as the Bellman equation, have been for much longer.

Recently, these advances have allowed us to showcase just how powerful reinforcement learning can be.

We’ve seen how

This is just a few years after the original AlphaGo already beat a world champion in Go.

We’ve seen real-world robots learn how to walk, and even recover after being kicked over, despite only being trained using simulation.

Simulation is nice because it doesn’t require actual hardware, which is expensive. If your agent falls down, no real damage is done.

We’ve seen real-world robots learn hand dexterity, which is no small feat.

Walking is one thing, but that involves coarse movements. Hand dexterity is complex - you have many degrees of freedom and many of the forces involved are extremely subtle.

Imagine using your foot to do something you usually do with your hand, and you immediately understand why this would be difficult.

Last but not least - video games.

Even just considering the past few months, we’ve seen some amazing developments. AIs are now beating professional players in

So what makes this course different from the first two?

Now that we know deep learning works with reinforcement learning, the question becomes: how do we improve these algorithms?

This course is going to show you a few different ways: including the powerful

Evolution strategies is a new and fresh take on reinforcement learning, that kind of throws away all the old theory in favor of a more "black box" approach, inspired by biological evolution.

What’s also great about this new course is the variety of environments we get to look at.

First, we’re going to look at the classic

Second, we’re going to look at

Finally, we’re going to look at

What do you get if you sign up for the VIP version of this course? A brand new

Thanks for reading, and I’ll see you in class!

Suggested Prerequisites:

- calculus
- matrix arithmetic (adding, multiplying)
- probability
- Object-oriented programming
- Python coding: if/else, loops, lists, dicts, sets
- Numpy coding: matrix and vector operations, loading a CSV file
- linear regression, logistic regression
- neural networks and backpropagation
- Know how to build a convolutional neural network (CNN) in TensorFlow
- Markov Decision Proccesses (MDPs), Q-Learning

- Introduction (03:46) (FREE preview available)
- Outline (07:46)
- Where to get the code (04:36)

- Review Section Introduction (04:01)
- The Explore-Exploit Dilemma (13:35)
- Markov Decision Processes (MDPs) (20:19)
- Monte Carlo Methods (07:54)
- Temporal Difference Learning (TD) (17:16)
- OpenAI Gym Warmup (06:46)
- Review Section Summary (07:28)

- A2C Section Introduction (07:54)
- A2C Theory (part 1) (20:40)
- A2C Theory (part 2) (06:47)
- A2C Theory (part 3) (03:13)
- A2C Demo (03:09)
- A2C Code - Rough Sketch (07:10)
- Multiple Processes (08:47)
- Environment Wrappers (11:49)
- Convolutional Neural Network (05:31)
- A2C (17:21)
- A2C Section Summary (06:40)

- DDPG Section Introduction (03:37)
- Deep Q-Learning (DQN) Review (09:20)
- DDPG Theory (18:42)
- MuJoCo (18:53)
- DDPG Code (part 1) (18:30)
- DDPG Code (part 2) (06:35)
- DDPG Section Summary (04:24)

- ES Section Introduction (06:25)
- ES Theory (20:16)
- Notes on Evolution Strategies (08:50)
- ES for Optimizing a Function (06:33)
- ES for Supervised Learning (06:39)
- Flappy Bird (12:09)
- ES for Flappy Bird in Code (15:01)
- ES for MuJoCo in Code (07:49)
- ES Section Summary (05:02)

- What is the Appendix? (02:48)
- Windows-Focused Environment Setup 2018 (20:21)
- How to install Numpy, Scipy, Matplotlib, Pandas, IPython, Theano, and TensorFlow (17:33)
- Is this for Beginners or Experts? Academic or Practical? Fast or slow-paced? (22:04)
- How to Code Yourself (part 1) (15:55)
- How to Code Yourself (part 2) (09:23)
- Proof that using Jupyter Notebook is the same as not using it (12:29)
- What order should I take your courses in? (part 1) (11:19)
- What order should I take your courses in? (part 2) (16:07)
- Python 2 vs Python 3 (04:38)
- How to Succeed in this Course (Long Version) (10:25)
- Is Theano Dead? (10:03)
- Where to get discount coupons and FREE deep learning material (02:21)

- TD3 (Twin Delayed DDPG) Theory
- TD3 Code