Skip to content

FrancescoMonaco/BiasMedical

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 

Repository files navigation

Debiasing of a ML model for managing the health of populations

The projects consists in taking the base model showed in "Obermeyer, Z., Powers, B., Vogeli, C., Mullainathan, S., Dissecting racial bias in an algorithm used to manage the health of populations, Science 366, https://escholarship.org/content/qt6h92v832/qt6h92v832.pdf, pp.447-453, 2019" and applying some methods for debiasing known in the literature. A comparison is then carried out between the models, with a correlated discussion on tradeoffs and accuracy. The models developed are:

  • Decorrelated model (FairLearn Decorrelator)
  • Augmented model (ADASYN)
  • Model with a new label
  • Adversarial neural network

Theory is explored in: https://thesis.unipd.it/handle/20.500.12608/33543

About

Compare various methods to debias a ML task (project for the AI course)

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published