Simula@BI: Explainable AI - from game theory with love
Speaker: Inga Strumke
The branch of artificial intelligence dedicated to learning from data, known as machine learning, has impressed us over the last decade by producing models that sometimes exceed human abilities. There's only one problem: We usually don't understand what such models have understood.
This seminar discusses the perspectives and motivations of three stakeholders' positions interested in explanations of machine learning models, before we explore an approach based on one of the crown jewels of cooperative game theory.
In addition to time for questions and discussions, there will be hands-on exercises - so bring a pen, paper and your game face!