Book Image

Mastering Reinforcement Learning with Python

By : Enes Bilgin
Book Image

Mastering Reinforcement Learning with Python

By: Enes Bilgin

Overview of this book

Reinforcement learning (RL) is a field of artificial intelligence (AI) used for creating self-learning autonomous agents. Building on a strong theoretical foundation, this book takes a practical approach and uses examples inspired by real-world industry problems to teach you about state-of-the-art RL. Starting with bandit problems, Markov decision processes, and dynamic programming, the book provides an in-depth review of the classical RL techniques, such as Monte Carlo methods and temporal-difference learning. After that, you will learn about deep Q-learning, policy gradient algorithms, actor-critic methods, model-based methods, and multi-agent reinforcement learning. Then, you'll be introduced to some of the key approaches behind the most successful RL implementations, such as domain randomization and curiosity-driven learning. As you advance, you’ll explore many novel algorithms with advanced implementations using modern Python libraries such as TensorFlow and Ray’s RLlib package. You’ll also find out how to implement RL in areas such as robotics, supply chain management, marketing, finance, smart cities, and cybersecurity while assessing the trade-offs between different approaches and avoiding common pitfalls. By the end of this book, you’ll have mastered how to train and deploy your own RL agents for solving RL problems.
Table of Contents (24 chapters)
Section 1: Reinforcement Learning Foundations
Section 2: Deep Reinforcement Learning
Section 3: Advanced Topics in RL
Section 4: Applications of RL

Challenges in meta-reinforcement learning

The main challenges regarding meta-RL, following (Rakelly, 2019), are as follows:

  • Meta-RL requires a meta-training step over various tasks, which are usually hand-crafted. A challenge here is to create an automated procedure to generate these tasks.
  • The exploration phase that is supposed to be learned during meta-training is in practice is not efficiently learned.
  • Meta-training involves sampling from an independent and identical distribution of tasks, which is not a realistic assumption. So, one goal is to make meta-RL more "online" by making it learn from a stream of tasks.

In addition to these challenges, it is important to note meta-RL methods will not work as well as the other methods, such as domain randomization, in complex tasks like robot hand manipulation. As the research in this area progresses, we can expect to see this gap to decrease and meta-RL make its way to mainstream with its unique advantages...