Reinforcement Learning Foundations

Course details

Innovations in finance, health, robotics, and a variety of other sectors have been made possible with reinforcement learning (RL), which involves the training of machines to learn from their environment. Many top tech companies are investing heavily in this field. In this course, instructor Khaulat Abdulhakeem helps you learn the basics of this relatively new, but valuable skill. Get to know the key terminology used in RL, how RL plays a major role in the advancement of AI, and the kinds of problems you can use RL to solve. Khaulat shows you how to define and represent reinforcement learning problems. She also delves into RL algorithms, including the Monte Carlo and temporal difference methods. Plus, she explores deep and multi-agent RL, as well as how inverse learning works and how it can help agents learn by imitation.

Concepts

Introduction

Reinforcement learning in a nutshell

Getting Started with Reinforcement Learning

Terms in reinforcement learning
A basic RL problem
Markov decision process
A basic RL solution

Reinforcement Learning Algorithms

Monte Carlo method
Temporal difference methods
Other RL algorithms

Monte Carlo Method

The setting
Exploration and exploitation
Monte Carlo prediction
First visit and every visit MC prediction
Monte Carlo control
Additional modifications

Temporal Difference Methods

The setting
SARSA
SARSAMAX (Q-learning)
Expected SARSA

Modified Forms of Reinforcement

Deep reinforcement learning
Multi-agent reinforcement learning
Inverse reinforcement learning

Conclusion

Your reinforcement learning journey