MIME: Minority Inclusion for Majority Group Enhancement of AI Performance
This addresses fairness and performance in AI for diverse stakeholders, though it appears incremental by challenging a specific assumption.
The paper tackles the misconception that minority inclusion in AI training data does not benefit majority groups, showing that including minority samples can improve test error for the majority group, with experimental results on six datasets.
Several papers have rightly included minority groups in artificial intelligence (AI) training data to improve test inference for minority groups and/or society-at-large. A society-at-large consists of both minority and majority stakeholders. A common misconception is that minority inclusion does not increase performance for majority groups alone. In this paper, we make the surprising finding that including minority samples can improve test error for the majority group. In other words, minority group inclusion leads to majority group enhancements (MIME) in performance. A theoretical existence proof of the MIME effect is presented and found to be consistent with experimental results on six different datasets. Project webpage: https://visual.ee.ucla.edu/mime.htm/