강의

멘토링

커뮤니티

AI Technology

/

Deep Learning & Machine Learning

Reinforcement learning made easy with R

We will learn about Q-learning and Deep Q-learning, and have time to implement reinforcement learning in R. We will cover the entire reinforcement learning content, from Deep Q-network to Self-imitation learning and Random Network Distillation.

(4.3) 3 reviews

96 learners

Level Intermediate

Course period Unlimited

  • coco
Machine Learning(ML)
Machine Learning(ML)
R
R
Reinforcement Learning(RL)
Reinforcement Learning(RL)
Machine Learning(ML)
Machine Learning(ML)
R
R
Reinforcement Learning(RL)
Reinforcement Learning(RL)

What you will gain after the course

  • Reinforcement learning theory

  • From Q-learning to Deep Reinforcement Learning

  • Several reinforcement learning techniques for exploration

🙆🏻‍♀ Beyond Q-learning and Deep Q-learning to RND🙆🏻‍♂

🗒 Course Introduction

The reinforcement learning boom started with AlphaGo. Did you know that reinforcement learning was an algorithm that existed long before AlphaGo?

Reinforcement learning is generally known as a field with a high barrier to entry. While the emergence of AlphaGo sparked interest, the complex nature of the subject makes it challenging to learn. For those who have been eager to learn reinforcement learning but have been hesitant to even begin, this course summarizes the key points. From Q-learning to DQN, and beyond DQN, this course explores the sparse reward problem, a key challenge in reinforcement learning, and various ideas for solving it. This course will provide a comprehensive overview of reinforcement learning in a short period of time.

🌈 What on earth is reinforcement learning?

We will explain step-by-step, using examples, what reinforcement learning is, what elements it contains, and how learning progresses.

🌈 Q-learning that you can solve by hand

Explaining it in words alone isn't enough. Let's try solving Q-learning problems ourselves to truly grasp the concepts of reinforcement learning.


🌈 DQN, the foundation of deep reinforcement learning

This book summarizes the core concepts of Deep Reinforcement Learning, from Deep Q-network (DQN) to various DQN variants including PerDQN, actorcritic, and Self-Imitation learning.

🌈 Sparse reward problem, the main problem of reinforcement learning

We will discuss the sparse reward problem, which is a major problem in reinforcement learning, and discuss various techniques to solve it.

We mainly talk about 'curiosity' or 'prediction error' and introduce several algorithms that utilize them.

(SIL, Random Network Distillation, etc.)

🌈 Implementing DQN/ActorCritic/SIL/RND directly in R

Unless you implement it yourself with code, you only know half of it, right? Let's write reinforcement learning algorithms for the most important models in R and examine the results together.

And let's also see if RND for Exploration really works.

🙋🏻‍♂️ Expected questions related to the lecture

Q. Do you have any player knowledge?
A. It would be good to have a basic understanding of machine learning and NN.

Q. Don't you do any practical training with Python?
A. Currently, I have uploaded the lecture by implementing the practical code in R, and I plan to upload the practical code in Python in the future. (I will upload it in addition to this lecture, rather than opening it as another lecture.)

Recommended for
these people

Who is this course right for?

  • For those who want to learn reinforcement learning easily

  • Anyone who wants to learn the entire reinforcement learning in a short period of time

Need to know before starting?

  • Intermediate R programming skills

  • Basic understanding of neural networks

  • Basic knowledge of machine learning

Hello
This is

8,390

Learners

509

Reviews

136

Answers

4.4

Rating

20

Courses

I am an unemployed scholar who majored in statistics as an undergraduate, earned a PhD in industrial engineering (artificial intelligence), and is still studying.

Awards ㆍ 6th Big Contest: Game User Churn Algorithm Development / NCSOFT Award (2018) ㆍ 5th Big Contest: Loan Delinquency Prediction Algorithm Development / Korea Association for ICT Promotion

Awards

ㆍ 6th Big Contest Game User Churn Prediction Algorithm Development / NCSOFT Award (2018)

ㆍ 5th Big Contest Loan Defaulter Prediction Algorithm Development / Korea Association for ICT Promotion (KAIT) Award (2017)

ㆍ 2016 Weather Big Data Contest / Korea Institute of Geoscience and Mineral Resources President's Award (2016)

ㆍ 4th Big Contest: Development of Insurance Fraud Prediction Algorithm / Finalist (2016)

ㆍ 3rd Big Contest Baseball Game Prediction Algorithm Development / Minister of Science, ICT and Future Planning Award (2015)

* blog : https://bluediary8.tistory.com

My primary research areas are data science, reinforcement learning, and deep learning.

I am currently doing crawling and text mining as a hobby :)

I developed an app called Marong that uses crawling to collect and display only popular community posts,

I also created a restaurant recommendation app by collecting lists of famous restaurants and blog posts from across the country :) (it failed miserably..)

I am currently a PhD student researching artificial intelligence.

I even developed a restaurant recommendation app by collecting blog posts and lists of top-rated restaurants across the country :) (though it failed miserably...) Now, I am a PhD student researching artificial intelligence.

I even developed a restaurant recommendation app by collecting lists of famous restaurants and blogs from all over the country :) (It failed miserably...) Now, I am a PhD student researching artificial intelligence.

I even developed a restaurant recommendation app by collecting lists of famous restaurants and blogs from all over the country :) (It failed miserably...) Now, I am a PhD student researching artificial intelligence.

I even developed a restaurant recommendation app by collecting lists of famous restaurants and blogs from all over the country :) (It failed miserably...) Now, I am a PhD student researching artificial intelligence.

Curriculum

All

20 lectures ∙ (4hr 31min)

Course Materials:

Lecture resources
Published: 
Last updated: 

Reviews

All

3 reviews

4.3

3 reviews

  • chihooi19851635님의 프로필 이미지
    chihooi19851635

    Reviews 5

    Average Rating 5.0

    5

    100% enrolled

    It helped me a lot

    • krstyle03v님의 프로필 이미지
      krstyle03v

      Reviews 5

      Average Rating 5.0

      5

      70% enrolled

      I tried to watch several reinforcement learning lectures, but I personally think that this person is the best at explaining the overall picture of reinforcement learning in an interesting way. I have only read up to section 2, but it keeps making me curious. However, since I have limited access to R, I have a hard time understanding the code, so I really hope that the Python code comes out soon.

      • doabc님의 프로필 이미지
        doabc

        Reviews 3

        Average Rating 4.0

        3

        35% enrolled

        It's a bit difficult to look at without a foundation in artificial intelligence + reinforcement learning.

        $42.90

        coco's other courses

        Check out other courses by the instructor!

        Similar courses

        Explore other courses in the same field!