Multi-group machine learning: theory and algorithms

The goal of this project is to advance the theory and algorithms for machine learning problems that require performance guarantees at subpopulation-level granularity.

• Simple and near-optimal algorithms for hidden stratification and multi-group learning
  • Christopher Tosh and Daniel Hsu
  • ICML 2022
• Distribution-specific auditing for subgroup fairness
  • Daniel Hsu, Jizhou Huang, and Brendan Juba
  • FORC 2024
• Multi-group learning for hierarchical groups
  • Samuel Deng and Daniel Hsu
  • ICML 2024
• Group-wise oracle-efficient algorithms for online multi-group learning
  • Samuel Deng, Daniel Hsu, and Jingwen Liu
  • NeurIPS 2024

Work on this project has been supported by the ONR under grant N00014-24-1-2700; by the NSF under grants CCF-1740833, IIS-1563785, and IIS-2040971; and by a JP Morgan Faculty Award.