Simula@BI: Should we stop using black box machine learning models and use interpretable models instead?

Speaker: Lars Henry Berge Olsen, University of Oslo

  • Starts:13:30, 31 March 2022
  • Ends:14:30, 31 March 2022
  • Location:Auditorium B2-060, BI Campus Oslo, or Stream
  • Contact:Siri Johnsen (siri.johnsen@bi.no)

In this talk, Lars Henry Berge Olsen is going to critically discuss the article "Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead" by Cynthia Rudin. Rudin is a renowned researcher in the field of interpretability and received the AAAI Squirrel AI Award for Artificial Intelligence for the Benefit of Humanity in 2021.

The article raises several important and interesting questions: what is an interpretable model? Can we properly explain a black box model? Who is responsible for incorrect predictions made by black box machine learning models? Do interpretable models always exist? All of these thought-provoking questions will be discussed in the talk.