Skills You’ll Learn
Understanding the big picture of recommendation systems
Understanding the development process and goals of a service called a recommendation system
How to evaluate recommendation systems
Online testing, offline testing and their methods
Rule-based, collaborative filtering-based recommendation system
Understand contextual analysis and algorithm application rather than just using packages
Recommended similar items based on embedded models
Recommended skills similar to embedding, such as Meta-Prod2Vec and graph deep learning
Using application applications
Use a vector math engine that utilizes latent factors, embedded results, etc.
✍ To improve business results through a recommendation system,
This has already been proven by many IT/commerce/platform companies.
▶ From understanding recommendation algorithms to how to implement recommendation systems based on various services
A service where the recommendation system is the core, Food Delivery's Yogiyo Machine Learning Engineer will fully help you get started with the recommendation system.
Is it enough to learn algorithms and theory?
Service-optimized problem solving
This is the beginning of the recommendation system.
In e-commerce and food delivery, I've experienced both areas where 'recommendations' play an important role in service. While working as a recommendation system engineer, I also developed recommendation algorithms suitable for each situation (context) that occurs in the service, and also developed a recommendation system based on personalization (personalization). (E-commerce service personalization/Delivery app service recommendation upgrade)
What I felt while building a recommendation system in the business It is more important to go beyond algorithms and theories to solve problems according to each business situationIt was.
Trying one line of a package
It's not a recommendation system.
When I search for materials to study recommendation systems, I often finish by running a line of package code after explaining a paper or theory of an algorithm in a verbose manner. However, simply knowing the theory and running a package line by line is not a recommendation system.
▶ Learn from an incumbent who has experienced both e-commerce and food delivery, where recommendations play an important role
To accurately understand the “situation” and “service” elements of “recommendation,” and to find solutions that match themThis is the process of properly establishing a recommendation system. Algorithms and theories can also be fully learned through papers, Business problem solving is an area that you can't understand at all unless you listen to business people.
Classes are held with recommended system engineers who have experienced both e-commerce and food delivery services so that they can take a practical perspective as it is. Starting with understanding algorithms, we will create a recommendation system tailored to the service and each business situation.
▶ Class curriculum
📌 Why was this algorithm used in this situation
📌 How can business problems be improved
In this class, students study recommendation systems with a focus on theory (30%) + practical code solving (70%). But it's not all theory or code. Rather, we'll focus on why these algorithms are used in this situation and how the problem can be improved.
Through this class, I will tell you all about what a recommendation system is, and how to define and solve the 'situation' and 'service' elements that I have learned through my experience with user services.
▶ Course Eligibility
▶ Course features
📧 1:1 coaching directly taught by Yamarae
You can ask two questions per coaching ticket.
- If you ask questions in as much detail as possible, we will be able to answer them more accurately.
1. Career consulting related to employment, turnover, and job transition (recommended)
- I'm looking directly at the resumes and interviews of job applicants at the company I'm currently working for.
- Please send us information including major/job-desired company/job-desired domain/job change/current career.
- We will respond in around 500 characters about the current situation and how to write a resume and design the required portfolio according to the desired direction.
2. Questions about code and analysis related to classes
- I will give you a supplementary explanation of the difficult content in the class in around 300 characters.
- We will fix the parts that cause errors during execution.
📌 How to use coaching tickets
- Click [My Classes] on the Class 101 web or app.
- Go to [My Class], go to [Coaching Ticket Mission] and click [Get Coaching].
- Please fill it out in [Write a post] and send it!
- Coaching is based on the date the question is received, and you will receive an answer within 7 to 10 days.
🚨 Coaching vouchers can be used for 20 weeks after the date of purchase, and there is no refund for unused use within the period.
- 📢 The package is subject to some changes, and we will be fully informed if there are any changes.
Curriculum
Creator
Yamarae
Hello, I'm the creator Yamarae.
As a short summary of my career, I worked in the e-commerce field for about a few years creating recommended/search systems, and now I'm working as an ML engineer for recommended/search systems at a place called Delivery Hero.
I like to give lectures and I like to write, so I do a lot of this and that outside of the company.
- Book <This Is Data Analysis with Python (Hanbit Media, 2020) >
- Running a machine learning blog (https://yamalab.tistory.com/)
- Fast Campus Python/Data Science Lectures (2017 to present)
- Learning Spoons Data Field Career Course (2021-present)
- Python-Advisory Committee of Korea Academy (2021)
- Many other lectures
I'm from a non-major. I spent most of my time in college studying graphic design and UX design. But then I started making something in Python, and then I also became interested in recommendation systems by chance. Eventually, at some point, I realized that the aspiring designer became an ML engineer creating a recommendation system.
That's why I really like teaching from a beginner's point of view. This is because people who don't know coding well know the pain of learning Python, and people who don't know the M character of ML know it.
I hope this class will help those who want to step into a new field.
Yamarae
