Selective Classification Can Magnify Disparities Across Groups
The Stanford AI Lab Blog
OCTOBER 13, 2021
Across a range of applications from vision 1 2 3 and NLP 4 5 , even simple selective classifiers, relying only on model logits, routinely and often dramatically improve accuracy by abstaining. As a motivating example, consider the task of diagnosing pleural effusion, or fluid in the lungs, from chest X-rays.
Let's personalize your content