The model's face in the mirror: fairness and transparency
You have to choose between fairness measures, because satisfying them all at once is mathematically impossible. You will see this as an equation rather than an opinion.
4 steps
225 XP
A free account is needed
Start the lesson →
Sources
- Chouldechova, A. 2017 · Fair Prediction with Disparate Impact: A Study of Bias in Recidivism Prediction Instruments · Big Data 5(2)
- Kleinberg, J. et al. 2017 · Inherent Trade-Offs in the Fair Determination of Risk Scores · ITCS 2017
- Hardt, M. et al. 2016 · Equality of Opportunity in Supervised Learning · NeurIPS 2016
- Mitchell, M. et al. 2019 · Model Cards for Model Reporting · FAT* 2019
ML Academy · an interactive machine learning course that runs in your browser ·
All lessons