The mythos of model interpretability

   page       BibTeX_logo.png   
Zachary C. Lipton
Communications of the ACM 61(10), pages 36–43
October 2018

In machine learning, the concept of interpretability is both important and slippery.