Advanced Reinforcement Learning in Python: from DQN to SAC

Description

Advanced Reinforcement Learning in Python: from DQN to SAC, the advanced reinforcement learning course in Python from DQN to SAC has been published by Udemy Academy. This is the most complete advanced reinforcement learning course on Udemy. In this course, you will learn some of the most powerful deep learning algorithms in Python using PyTorch and PyTorch lightning. From scratch, you will implement adaptive algorithms that solve control tasks based on experience. You will learn to combine these techniques with neural networks and deep learning methods to create adaptive AI agents capable of solving decision-making tasks.

This course introduces you to the most advanced reinforcement learning techniques and also prepares you for the next courses in this series that examine other advanced methods that excel in other tasks. This course focuses on developing practical skills; Therefore, after learning the most important concepts of each family of methods, you implement one or more algorithms from scratch in jupyter notebooks.

Leveling modules:

Reminder: Markov Decision Process (MDP)
Reminder: Q-Learning
Reminder: A brief introduction to neural networks
Reminder: Deep Q-Learning
Reminder: Policy gradient methods

Advanced reinforcement learning:

PyTorch Lightning
Hyperparameter adjustment with optona
Deep learning for continuous action spaces (Normalized Advantage Function – NAF)
Policy gradient Deep Determinant (DDPG)
Delayed Conjugate DDPG (TD3)
Simple actor-critic
Hindsight Experience Replay (HER)

What you will learn

You will get to know some of the most advanced reinforcement learning algorithms
You will learn how to build artificial intelligence that can operate in a complex environment to achieve its goals.
Create advanced reinforcement learning agents using the most popular Python tools
Learn how to perform hyperparameter tuning (determining the best lab conditions for artificial intelligence learning)
Understand the basic learning process of each algorithm.
Debugging and development of presented algorithms.
Understanding and implementing new algorithms from research articles

Who is this course suitable for?

Developers who want to work in machine learning
Data scientists and ML professionals looking to expand their knowledge base.
Robotics students and researchers
Engineering students and researchers