Skip to content

v0.1.0 release

Latest
Compare
Choose a tag to compare
@k-narita0209 k-narita0209 released this 09 Jun 05:59
· 49 commits to main since this release
5022910

v0.1.0

Intersectional Fairness 0.1.0 Release Notes


Description

Intersectional Fairness (ISF) is a bias detection and mitigation technology for intersectional bias, which combinations of multiple protected attributes causes.

ISF leverages the existing single-attribute bias mitigation methods to make a machine-learning model fair regarding intersectional bias.

Approaches applicable to ISF are pre-, in-, and post-processing. For now, ISF supports Adversarial Debiasing, Equalized Odds, Massaging, and Reject Option Classification.

Highlights

  • Python 3.7, AIF360 0.4 now supported
  • Datasets Interface
  • Bias mitigation Algorithms
    • Adversarial Debiasing
    • Equalized Odds
    • Massaging
    • Reject Option Classification
  • Fairness metrics
    • Demographic Parity
    • Equal Opportunity
    • Equalized Odds
    • F1Parity
  • Accuracy metrics
    • Balanced Accuracy
    • F1
  • Related techniques used in tutorials
    • Logistic Regression
    • Disparate Impact