Fairness-Aware Learning Systems

Earlier work on fairness metrics and algorithmic interventions.

Summary

This project collects my earlier work on fairness-aware machine learning and the engineering work around fairness methods.

Research Question

How should fairness criteria be represented and operationalized in learning systems whose outputs affect people asymmetrically?

Methodology

  • Fairness metrics for survival analysis
  • In-processing algorithms during Julia Summer of Code
  • Practical interface design for fairness methods in ML tooling

Key Contributions

  • Contributed to fairness metrics for time-to-event modeling
  • Implemented fairness in-processing algorithms in the MLJ ecosystem
  • Established an early line of work on responsible ML systems

Outputs

References

2022

  1. adjacent interests
    Flexible Group Fairness Metrics for Survival Analysis
    Raphael Sonabend, Florian Pfisterer, Alan Mishler, and 4 more authors
    In DSHealth, 2022
    Bib Paper