Organized by Ho-min Park (Ph.D. Student, Ghent University)
Welcome to the GUGC AI Summer School 2023! We're excited to host this intensive three-week program from June 1st to June 20th, providing students with hands-on experience in the rapidly evolving field of artificial intelligence (AI). Our mission is to empower the next generation of AI enthusiasts with the skills, knowledge, and tools necessary to tackle real-world challenges and drive innovation.
The program is designed for students with varying levels of AI expertise and is divided into two groups:
-
Beginner Group: Ideal for students new to AI and data science, this group will focus on foundational AI concepts, Python programming, and data analysis. Participants will apply their knowledge to real-world datasets, creating machine learning models and sharing their results in interactive Colab notebooks.
-
Advanced Group: Tailored for students with prior experience in data science, this group will explore cutting-edge AI topics and mentor beginner students. Advanced participants will also prepare and present lecture sessions for the entire summer school cohort, showcasing their expertise.
The beginner group had a total of 24 participants, out of which 4 were selected through a qualification test consisting of 30 questions related to basic mathematics and Python programming.
From June 15th, after the final exams, to July 10th, the start of the AISS, students are required to take more than 20 hours of online lectures. The online lectures cover essential subjects like linear algebra and calculus, which are crucial for understanding machine learning, as well as machine learning, computer architecture, and computer networks. You can find detailed lectures at the following URL:
The material learned will be validated through a quiz after the Week 1 orientation. The quiz consists of 60 questions that test both theory and application and 80 questions that require definitions of terms. All participants passed with scores of at least 70.
The Advanced group students prepared and graded the lectures and quiz questions for Weeks 1-2.
The Beginner group participants attend a two-week on-site lecture. The program is outlined below. Through the program, students reviewed the pre-training content and ultimately validated it through the final exam, consisting of 27 questions. All beginner participants passed with scores of 80 or above.
The Advanced group took charge of at least one session, conducted lectures or practical exercises, prepared exam questions, and graded them. The names in parentheses in the table below are the names of the Advanced group students who were in charge of the respective sessions.
| Day | Date | Morning | Afternoon |
|---|---|---|---|
| Mon | July 10th | Orientation | Quiz for checking homework |
| Tue | July 11th | Computer network and structure | Environment setting and Python review |
| Wed | July 12th | Data mining & machine learning | Machine learning practical (1) |
| Thu | July 13th | Interpretability methods (Yejin Lee) | Machine learning practical (2) (Yujin Kim) |
| Fri | July 14th | Artificial Neural Network | ANN practical with NumPy (Dongin Moon) |
| Mon | July 17th | Convolutional Neural Network (Jongbum Won) | CNN Practical (Jongbum Won) |
| Tue | July 18th | Evaluation, loss, and optimizations | CNN practical + (MinJae Chung) |
| Wed | July 19th | RNN and Transformer (Ganghyun Kim) | Transformer Practical (Ganghyun Kim) |
| Thu | July 20th | Un- and self-supervised learning (Jongbum Won) | Reinforcement Learning (Ganghyun Kim) |
| Fri | July 21st | Generative Adversarial Network | Final exam |
Afterward, all participants conducted a more in-depth exploration of self-selected topics. Below is a summary of their projects, including links to their presentations and Colab notebooks.
The Beginner group consists of students who are new to the field of data science. During their studies, they were tasked with applying their knowledge to four interesting data sets. They analyzed the datasets and performed all the processes of predicting through machine learning models, and wrote a Colab, an interactive Python tool for sharing this process and results. The following are the results of their presentations and the information about the Colab.
| Name | Dataset name | Presentation | Kaggle code |
|---|---|---|---|
| Dahee Kim | Exploring Breast Cancer Patterns | Link | Code link |
| Taekeun Kim | Loan Default Prediction | Link | Code link |
| Yujin Sung | Detecting Heart Diseases | Link | Code link |
| Jihun Kwon | Mushrooms: Distinguishing the toxicity | Link | Code link |
The Advanced group consists of students who have prior experience with data science. They were asked to prepare and present a lecture session for the Beginner group during the AI Winter School. During the self-study period, they explored and presented topics of their own interest.
The AI Summer School was a great success, and all participants gained valuable experience working on real-world AI projects. We hope that the work presented here can be useful for future researchers and developers.
We would like to express our gratitude to the Center for Biosystems and Biotech Data Science for providing the funding for this program. Additionally, we would like to extend our appreciation to the Student Intensive Research Training Program (IRTP) in the Academic Affairs team.