6.19 LR in R: Predicting Recidvism (3)

  • Background story by ProPublica: Machine Bias
  • Replication and extension by Dressel and Farid (2018): The Accuracy, Fairness, and Limits of Predicting Recidivism
    • Abstract: “Algorithms for predicting recidivism are commonly used to assess a criminal defendant’s likelihood of committing a crime. […] used in pretrial, parole, and sentencing decisions. […] We show, however, that the widely used commercial risk assessment software COMPAS is no more accurate or fair than predictions made by people with little or no criminal justice expertise. In addition, despite COMPAS’s collection of 137 features, the same accuracy can be achieved with a simple linear classifier with only two features.”
  • Very nice lab by Lee, Du, and Guerzhoy (2020): Auditing the COMPAS Score: Predictive Modeling and Algorithmic Fairness
  • We will work with the corresponding data and use it to grasp various concepts underlying statistical/machine learning

References

Dressel, Julia, and Hany Farid. 2018. “The Accuracy, Fairness, and Limits of Predicting Recidivism.” Sci Adv 4 (1): eaao5580.

Lee, Claire S, Jeremy Du, and Michael Guerzhoy. 2020. “Auditing the COMPAS Recidivism Risk Assessment Tool: Predictive Modelling and Algorithmic Fairness in CS1.” In Proceedings of the 2020 ACM Conference on Innovation and Technology in Computer Science Education, 535–36. ITiCSE ’20. New York, NY, USA: Association for Computing Machinery.