Chevron Left
返回到 Prediction and Control with Function Approximation

学生对 阿尔伯塔大学 提供的 Prediction and Control with Function Approximation 的评价和反馈

4.8
724 个评分

课程概述

In this course, you will learn how to solve problems with large, high-dimensional, and potentially infinite state spaces. You will see that estimating value functions can be cast as a supervised learning problem---function approximation---allowing you to build agents that carefully balance generalization and discrimination in order to maximize reward. We will begin this journey by investigating how our policy evaluation or prediction methods like Monte Carlo and TD can be extended to the function approximation setting. You will learn about feature construction techniques for RL, and representation learning via neural networks and backprop. We conclude this course with a deep-dive into policy gradient methods; a way to learn policies directly without learning a value function. In this course you will solve two continuous-state control tasks and investigate the benefits of policy gradient methods in a continuous-action environment. Prerequisites: This course strongly builds on the fundamentals of Courses 1 and 2, and learners should have completed these before starting this course. Learners should also be comfortable with probabilities & expectations, basic linear algebra, basic calculus, Python 3.0 (at least 1 year), and implementing algorithms from pseudocode. By the end of this course, you will be able to: -Understand how to use supervised learning approaches to approximate value functions -Understand objectives for prediction (value estimation) under function approximation -Implement TD with function approximation (state aggregation), on an environment with an infinite state space (continuous state space) -Understand fixed basis and neural network approaches to feature construction -Implement TD with neural network function approximation in a continuous state environment -Understand new difficulties in exploration when moving to function approximation -Contrast discounted problem formulations for control versus an average reward problem formulation -Implement expected Sarsa and Q-learning with function approximation on a continuous state control task -Understand objectives for directly estimating policies (policy gradient objectives) -Implement a policy gradient method (called Actor-Critic) on a discrete state environment...

热门审阅

WP

Apr 11, 2020

Difficult but excellent and impressing. Human being is incredible creating such ideas. This course shows a way to the state when all such ingenious ideas will be created by self learning algorithms.

AC

Dec 1, 2019

Well peaced and thoughtfully explained course. Highly recommended for anyone willing to set solid grounding in Reinforcement Learning. Thank you Coursera and Univ. of Alberta for the masterclass.

筛选依据:

51 - Prediction and Control with Function Approximation 的 75 个评论(共 131 个)

创建者 Andrew G

Jan 26, 2020

Did a good job of attaching a programming assignment to each lesson and giving clear and detailed instructions throughout

创建者 Alexander P

Dec 14, 2019

Great course on more advanced reinforcement learning techniques. Can't wait to apply these new skills in the wild.

创建者 Mathew

Jun 7, 2020

Very well structured and a great compliment to the Reinforcement Learning (2nd Edition) book by Sutton and Barto.

创建者 Ayan S

Jul 4, 2021

I really liked the lectures and how they clearly explained all the necessary details of such difficult topic.

创建者 Johannes

Sep 13, 2021

T​his course is as excellent as its predecessors! Well-structured, engaging and with clear explanations.

创建者 Joosung M

Jun 14, 2020

The course materials were very informative, the assignments were challenging enough. Highly recommended!

创建者 Tolga K

Dec 25, 2020

Great course, great material and notebooks like previous courses. It was a great experience. Thank you!

创建者 J B

Oct 13, 2020

Very helpful course. Excellent delivery and practical labs. There's even someone helping in the forum!

创建者 Eduardo I L H

Jan 14, 2021

Excellent course. Focused in the theory of function approximation for reinforcement learning.

创建者 Yitao H

Aug 29, 2021

Intellectually challenging experience to combine supervised learning into RL framework!

创建者 Huang C

Jan 25, 2022

Great course to take for combining function approximations with reinforcement learning

创建者 RICARDO A F S

Nov 21, 2020

A great course, I took a long time doing the assignments, but in the end I solved it

创建者 Artur M

Nov 3, 2020

Great course! Wished to see more about policy gradient methods, but it was awesome.

创建者 George M

Mar 11, 2021

Comprehensive and intensive course.

More challenging than the previous two courses.

创建者 Chang, W C

Oct 14, 2019

The course presentation is wonderful. I can't stop after I watch the first video.

创建者 Rishi R

Aug 3, 2020

It has amazing content with no compromise on concepts yet holds simplicity.

创建者 Kaustubh S

Dec 24, 2019

It was a wonderful course. To the point yet well-explained concepts.

创建者 Max C

Nov 1, 2019

I had a much better experience with the autograder than in course 2.

创建者 Sergey M

Oct 15, 2021

Very nice and helpful course, very well organized and explained.

创建者 LIWANGZHI

Jan 27, 2020

Everything is amazing in this course! Dont miss it!

创建者 Pachi C

Dec 31, 2019

Fantastic course and great content and teachers!!!

创建者 석박통합김한준

Apr 25, 2020

Excellent course! Never be replaced! Thank you!

创建者 Raktim P

Dec 17, 2019

Great Course! Highly recommended for beginners.

创建者 Ola D

Jun 15, 2022

F​antastic course with fantastic instructors

创建者 İbrahim Y

Oct 5, 2020

the course is the intro for high level RL