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
726 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:

126 - Prediction and Control with Function Approximation 的 133 个评论(共 133 个)

创建者 Charles X

Jun 21, 2021

Gets hard to understand.

创建者 Quarup B

Jul 25, 2021

Content is great, but the text is super dense -- slow read for me. The lectures are much clearer, although also a bit dense / quick paced to retain the information long term (especially if one wishes to skip the reading).

创建者 Prashant M

Jun 7, 2020

great course material but you need read the RL book through out the course. Also assignments are bit difficult, oops concept is mandatory.

创建者 Justin N

Mar 31, 2020

Lectures are pretty good, but the programming exercises are extremely easy. All of the problems are rather contrived as well.

创建者 Yassine B

May 4, 2020

I think It must be more deep neural networks dedicated course and not focus on coarse and tile coding!!!

创建者 Bernard C

May 24, 2020

Course was good, but assignments were not well constructed. Problems with the unit tests were frequent.

创建者 Lars R

Aug 23, 2021

Feels to be too focussed on theory and math, instead on practically applying the best techniques.

创建者 Vasilis V

Jul 11, 2020

Needs more work in my opinion. It's not bad of course. I just believe that more intuition should be built with better examples, outside the text book rather than going through the actual mathematical proofs