Topics in machine learning for education

Content

Computer-based learning environments such as intelligent tutoring systems, educational games, and interactive simulations yield large amounts of data, which can be analyzed to gain new insights into human learning. Traditionally, students’ learning behavior and strategies has primarily been studied by learning scientist using instruments such as surveys or one-to-one interviews. The data sets collected by digital education allow for performing this research at scale, requiring computational approaches to thinking and analyzing about them.
The course is held as an advanced seminar, where original research papers have to be critically reviewed, presented, and discussed. Every week, we will focus on one research topic (read research paper and sometimes additional complementary materials). All students will write a short summary and review of the respective paper(s), and one student will present the paper in the class. The students will take turns in presenting the papers and leading the discussion throughout the semester.

Computer-based learning environments furthermore provide the opportunity to also offer individualization at scale: the users of educational systems are often very heterogeneous and therefore it is important to adapt to their specific needs and preferences. 

The goal of this seminar is to get familiar with the fundamental questions, issues, and core techniques in the fields of machine learning and artificial intelligence in education. The course is held as an advanced seminar, where original research papers have to be critically reviewed, presented, and discussed. Every week, we will focus on one research topic (read research paper(s) and sometimes additional complementary materials). All students will write a short summary and review of the respective paper(s), and one student will present the paper in the class.

Readings & Presentations

It is absolutely important to read the papers prior to attending class, because the course proceeds in the form of a discussion among the participants.

Attendance

Submission System

Grading

Logistics

Lecture:        Mondays, 10.15am-12.00pm

Instructor:     Tanja Käser

Room:           INF 220

Email:            tanja.kaeser (at) epfl.ch

Course attendance is mandatory, the course will be taught hybrid.
You can attend remotely through the following Zoom link:

Zoom link: https://epfl.zoom.us/j/64971198727
Password: mled21

Analysis & prediction of behavior

-

The presentation regarding the logistics of the course can be downloaded here and the presentation on knowledge tracing can be found here.

The presentation from week 2 can be downloaded from here.

The presentation from week 3 can be downloaded here.

The slides from this week's presentation can be downloaded here.

Additional article on fairness of recommenders for MOOC mentioned in class can be found here.

The slides from this week's presentation can be downloaded from here.

The slides from this week's presentation can be found here.

The slides from this week's presentation can be downloaded from here.

Here are some example guidelines: EDM, LAK, and CHI.

The slides from this week's presentation can be found here.

The presentation on modeling students' procrastination can be downloaded here.

The slides from this week's presentation can be downloaded from here.

The slides from this week's presentation can be downloaded as a pdf or pptx.