User Guide

• Fairness in Machine Learning
  • Fairness of AI systems
  • Types of harms
  • Concept glossary
    • Construct validity
  • Fairness assessment and unfairness mitigation
    • Group fairness, sensitive features
    • Parity constraints
    • Disparity metrics, group metrics
  • What traps can we fall into when modeling a social problem?
    • The Solutionism Trap
    • The Ripple Effect Trap
    • The Formalism Trap
    • The Portability Trap
    • The Framing Trap
  • References
• Assessment
  • Introduction
    • Identify types of harms
    • Identify the groups that might be harmed
    • Quantify harms
    • Compare quantified harms across the groups
  • Disaggregated metrics
  • Disaggregated metrics using MetricFrame
  • Multiple metrics in a single MetricFrame
    • Non-sample parameters
    • Multiclass metrics
    • Multiple sensitive features
    • Scalar results from MetricFrame
    • Control features for grouped metrics
  • Plotting
    • Plotting grouped metrics
  • Fairlearn dashboard
• Mitigation
  • Preprocessing
    • Correlation Remover
  • Postprocessing
  • Reductions
    • Fairness constraints for binary classification
    • Fairness constraints for multiclass classification
    • Fairness constraints for regression
    • Exponentiated Gradient
    • Grid Search
    • References
• Datasets
  • Adult Census Dataset
    • ACSIncome
  • Revisiting the Boston Housing Dataset
    • Introduction
    • Dataset Origin and Use
    • Dataset Issues
    • Fairness-related harms assessment
    • Discussion
• Installation and version guide
  • Installation Guide
    • Installation
    • Dependencies
  • Version guide
    • v0.1
    • v0.2.0
    • v0.3.0
    • v0.4.0
    • v0.4.1
    • v0.4.2
    • v0.4.3
    • v0.4.4
    • v0.4.5
    • v0.4.6
    • v0.5.0
    • v0.6.0
    • v0.6.1
    • v0.6.2
    • v0.7.0
    • v0.8.0
• Further Resources
  • Books
  • Papers
  • Online demos, talks, tutorials