Monthly Subscription Starting at AED 99 + VAT

A Complete Reinforcement Learning System (Capstone)
  • Level Professional
  • Duration 16 hours
  • Course by University of Alberta
  • Offered by

About

In this final course, you will put together your knowledge from Courses 1, 2 and 3 to implement a complete RL solution to a problem. This capstone will let you see how each component---problem formulation, algorithm selection, parameter selection and representation design---fits together into a complete solution, and how to make appropriate choices when deploying RL in the real world. This project will require you to implement both the environment to stimulate your problem, and a control agent with Neural Network function approximation. In addition, you will conduct a scientific study of your learning system to develop your ability to assess the robustness of RL agents. To use RL in the real world, it is critical to (a) appropriately formalize the problem as an MDP, (b) select appropriate algorithms, (c ) identify what choices in your implementation will have large impacts on performance and (d) validate the expected behaviour of your algorithms. This capstone is valuable for anyone who is planning on using RL to solve real problems. To be successful in this course, you will need to have completed Courses 1, 2, and 3 of this Specialization or the equivalent. By the end of this course, you will be able to: Complete an RL solution to a problem, starting from problem formulation, appropriate algorithm selection and implementation and empirical study into the effectiveness of the solution.

Modules

Course Introduction

    • 2 Videos
    • 2 Readings
    • 1 Discussion
Show info about module content

1 Discussions

  • Meet and Greet

2 Videos

  • Course 4 Introduction
  • Meet your instructors!

2 Readings

  • Reinforcement Learning Textbook
  • Pre-requisites and Learning Objectives

Final Project: Milestone 1

    • 2 Videos
Show info about module content

2 Videos

  • Initial Project Meeting with Martha: Formalizing the Problem
  • Andy Barto on What are Eligibility Traces and Why are they so named?

Project Resources

    • 2 Videos
    • 1 Programming
Show info about module content

1 Programming

  • MoonShot Technologies

2 Videos

  • Let's Review: Markov Decision Processes
  • Let's Review: Examples of Episodic and Continuing Tasks

Weekly Learning Goals

    • 1 Videos
Show info about module content

1 Videos

  • Meeting with Niko: Choosing the Learning Algorithm

Project Resources

    • 6 Videos
    • 1 Assignment
Show info about module content

1 Assignment

  • Choosing the Right Algorithm

6 Videos

  • Let's Review: Expected Sarsa
  • Let's Review: What is Q-learning?
  • Let's Review: Average Reward- A New Way of Formulating Control Problems
  • Let's Review: Actor-Critic Algorithm
  • Csaba Szepesvari on Problem Landscape
  • Andy and Rich: Advice for Students

Weekly Learning Goals

    • 1 Videos
Show info about module content

1 Videos

  • Agent Architecture Meeting with Martha: Overview of Design Choices

Project Resources

    • 3 Videos
    • 1 Assignment
Show info about module content

1 Assignment

  • Impact of Parameter Choices in RL

3 Videos

  • Let's Review: Non-linear Approximation with Neural Networks
  • Drew Bagnell on System ID + Optimal Control
  • Susan Murphy on RL in Mobile Health

Weekly Learning Goals

    • 1 Videos
Show info about module content

1 Videos

  • Meeting with Adam: Getting the Agent Details Right

Project Resources

    • 5 Videos
    • 1 Programming
Show info about module content

1 Programming

  • Implement Your Agent

5 Videos

  • Let's Review: Optimization Strategies for NNs
  • Let's Review: Expected Sarsa with Function Approximation
  • Let's Review: Dyna & Q-learning in a Simple Maze
  • Meeting with Martha: In-depth on Experience Replay
  • Martin Riedmiller on The 'Collect and Infer' framework for data-efficient RL

Weekly Learning Goals

    • 1 Videos
Show info about module content

1 Videos

  • Meeting with Adam: Parameter Studies in RL

Project Resources

    • 2 Videos
    • 1 Programming
Show info about module content

1 Programming

  • Completing the parameter study

2 Videos

  • Let's Review: Comparing TD and Monte Carlo
  • Joelle Pineau about RL that Matters

Congratulations!

    • 3 Videos
Show info about module content

3 Videos

  • Meeting with Martha: Discussing Your Results
  • Course Wrap-up
  • Specialization Wrap-up

Auto Summary

This comprehensive capstone course in Data Science & AI, led by Coursera, is designed for professionals aiming to deploy Reinforcement Learning (RL) solutions. Over 960 minutes, you'll integrate knowledge from prior courses to tackle real-world problems through problem formulation, algorithm selection, and empirical studies. Ideal for those with foundational expertise, the course emphasizes practical application and robustness of RL agents. Subscription options include Starter and Professional tiers.

Martha White
Instructors

Martha White

Adam White
Instructors

Adam White