Repository logo
 

On the Relation between Distributionally Robust Optimization and Data Curation (Student Abstract)

Accepted version
Peer-reviewed

Type

Article

Change log

Authors

Słowik, A 
Bottou, L 
Holden, SB 

Abstract

jats:pMachine learning systems based on minimizing average error have been shown to perform inconsistently across notable subsets of the data, which is not exposed by a low average error for the entire dataset. In consequential social and economic applications, where data represent people, this can lead to discrimination of underrepresented gender and ethnic groups. Distributionally Robust Optimization (DRO) seemingly addresses this problem by minimizing the worst expected risk across subpopulations. We establish theoretical results that clarify the relation between DRO and the optimization of the same loss averaged on an adequately weighted training dataset. A practical implication of our results is that neither DRO nor curating the training set should be construed as a complete solution for bias mitigation.</jats:p>

Description

Keywords

46 Information and Computing Sciences, 4611 Machine Learning, Machine Learning and Artificial Intelligence, Health Disparities

Journal Title

Proceedings of the 36th AAAI Conference on Artificial Intelligence, AAAI 2022

Conference Name

Journal ISSN

2159-5399
2374-3468

Volume Title

36

Publisher

Association for the Advancement of Artificial Intelligence (AAAI)