Machine Learning Skills with bite-size Workshops

Schliessen Icon

We create hands-on workshops for all BSc, MSc and PhD students from ETH to experience and understand recent developments in machine learning (ML). From large-language models to computer vision, our aim is to give students a basic understanding of the main technologies based on ML. The teaching material will be opened for all departments and lecturers to reproduce freely.

A basic understanding of recent developments in machine-learning (ML) and what some call artificial intelligence (AI) is essential for all ETH students as these developments affect every discipline. Based on students’ demand and feedback, we created a successful prototype workshop on large language models. We will now create a coherent curriculum of 5 to 6 hands-on workshops teaching the basis of ML and its applications. Our aim is to give students a basic understanding of the main technologies involved in  ML, from large-language models to computer vision. We will also raise awareness about current challenges in the field of  ML such as inherent gender biases or lock-in effects of proprietary ML models. Our pedagogical approach relies on project-based learning, peer learning and intrinsic motivation of students. We wish for all students to experience and understand recent developments in ML, focusing on how to use ML as a tool and work with it rather than on in-depth technical understanding. The workshops will be open to all ETH students regardless of study level or department  and most workshops will be accessible to students without any coding experience. The teaching material will be open source and made available for all departments to reproduce freely.

Machine Learning Workshop Principles

Our project is based on three principles that have proven their efficiency in the literature, at ETH and at the Student Project House: a. project-based learning, b. peer learning and c. intrinsic motivation.

  • Each workshop is conceived as a small project-based learning module. After a very short theoretical introduction (15 to 25 minutes) the participants jump straight into practice. We give them a challenge that they have to solve. For example, for our prototype workshop, participants have to make a chatbot that can answer questions about the Student Project House based on information on the SPH website. Special care is taken when selecting realistic and practical problems. We use Google Collab as a tool to allow students to have a well-documented code base they can work with even if they do not have prior coding skills.
  • Our workshops incorporate peer learning in two ways. First, to solve the challenge, students will be encouraged to collaborate and help each other. Second, workshops are be given by a member of the SPH staff helped by one or two voluntary students from our Digital Makerspace Manager pool. Their role is to interact with participants on an individual basis, answer their questions and help them to solve the challenge. These Digital Makerspace Managers also help create the workshops in order to bring a student perspective.
  • These workshops run purely on intrinsic motivation of students: no grades, no credits, no pressure, which has shown great results at SPH.
Question:
What effect did the innovative elements have on student learning?
Answer:
From the first workshop we notice very high student engagement during the workshop. All students engaged with the teaching material and some stayed long after the planned time to code away. Some students used the teaching material as a starting point to create a solution to their own problems.
Question:
How do you ensure (continuous) feedback on student learning and satisfaction?
Answer:
Shortly before the end of the workshop we ask students to fill in a feedback questionnaire that they access through a QR code. With this approach we can integrate the feedback into the next workshop. We ask 3 main questions:
- From 1 to 10 how likely is it that you would recommend this to a friend or colleague? This allows us to assess satisfaction through a net promoter score.
- What did you like?
- What can we improve? 
We use the same measurement method for all workshops at the Student Project House.
Question:
Which elements of your project would you recommend to others?
Answer:
- Project based learning and learning by doing
- Having a fellow student acting as a mentor and giving individual feedback during the sessions.
- Having challenges that are useful to students and that they can relate to.

loading