Train the Trainer: AI for the Common Good

A variety of topics will be covered, including but not limited to:

• Search algorithms
• Neural networks
• Causality
• Bayesian inference
• Handling uncertainty
• Handling small data sets
• Declarative languages
• Planning
• Recommender systems
• Time series prediction
• Reinforcement learning

The core pedagogical approach is centered around so-called “AI missions”, which all have the same structure.

– 9:30 A real though simplified problem is given to the students;

– 10:00 Participants tackle the problem, in pairs, on their own;

– 11:00 The tutor helps students move forward, and identifies problems;

– 12:30 Encountered problems are discussed in group.

– 13:00 Lunch break

– 14:00 The AI solution is presented; students try to unravel how it works.

– 14:30 The mechanisms behind the AI technique/approach are explained.

– 15:30 Small exercises are done to empower the participants.

– 16:30 Wrap-up and further reading

During the day, participants are continuously challenged by asking question, for example on ethical implications. The goal is to increase their critical thinking skills, while stimulating their creativity in how the techniques could be applied.

The second day looks as follows. Before noon, we deepen the knowledge:

– 09:30 Recap of the problem, with explanation of how it was constructed and how to adapt it to a new context, how to create new exercises;

– 11:30 Each trainer invents a new, simplified training exercise.

In the afternoon, we focus on training hands-on teaching skills.

– 13:00 In teams of 2 persons, one person trains the other using the self-invented exercise. The trainee analyses the teaching style, and notes down fortes and potential improvements for the trainer.

– 14:30 The participants of each team switch roles. At the end of this second session, they discuss each other’s strengths & potential for improvement.

– 16:00 Group discussion where the exercises are discussed (and why they were helpful in teaching about the technique + how they could be improved), as well as what they have learnt from the fortes and potential improvements.