KANDINSKY Patterns: A Swiss-Knife for the Study of Explainable AI
by Andreas Holzinger (Medical University Graz and Alberta Machine Intelligence Institute Edmonton, Canada), Peter Kieseberg (University of Applied Sciences, St.Poelten) and Heimo Mülller (Medical University Graz)
Kandinsky Patterns are mathematically describable, simple, self-contained, hence controllable test datasets for the development, validation and training of explainability in artificial intelligence (AI) and machine learning (ML).