Key points are not available for this paper at this time.
In machine learning, the concept of interpretability is both important and slippery.
Share your take
Add a clinician perspective alongside expert commentary.
Zachary C. Lipton (2018) studied this question.