Fairness Evaluation in Presence of Biased Noisy Labels

Riccardo Fogliato, Alexandra Chouldechova, Max G’Sell
Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics, PMLR 108:2325-2336, 2020.

Abstract

Risk assessment tools are widely used around the country to inform decision making within the criminal justice system. Recently, considerable attention has been devoted to the question of whether such tools may suffer from racial bias. In this type of assessment, a fundamental issue is that the training and evaluation of the model is based on a variable (arrest) that may represent a noisy version of an unobserved outcome of more central interest (offense). We propose a sensitivity analysis framework for assessing how assumptions on the noise across groups affect the predictive bias properties of the risk assessment model as a predictor of reoffense. Our experimental results on two real world criminal justice data sets demonstrate how even small biases in the observed labels may call into question the conclusions of an analysis based on the noisy outcome.

Cite this Paper


BibTeX
@InProceedings{pmlr-v108-fogliato20a, title = {Fairness Evaluation in Presence of Biased Noisy Labels}, author = {Fogliato, Riccardo and Chouldechova, Alexandra and G'Sell, Max}, booktitle = {Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics}, pages = {2325--2336}, year = {2020}, editor = {Silvia Chiappa and Roberto Calandra}, volume = {108}, series = {Proceedings of Machine Learning Research}, month = {26--28 Aug}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v108/fogliato20a/fogliato20a.pdf}, url = { http://proceedings.mlr.press/v108/fogliato20a.html }, abstract = {Risk assessment tools are widely used around the country to inform decision making within the criminal justice system. Recently, considerable attention has been devoted to the question of whether such tools may suffer from racial bias. In this type of assessment, a fundamental issue is that the training and evaluation of the model is based on a variable (arrest) that may represent a noisy version of an unobserved outcome of more central interest (offense). We propose a sensitivity analysis framework for assessing how assumptions on the noise across groups affect the predictive bias properties of the risk assessment model as a predictor of reoffense. Our experimental results on two real world criminal justice data sets demonstrate how even small biases in the observed labels may call into question the conclusions of an analysis based on the noisy outcome. } }
Endnote
%0 Conference Paper %T Fairness Evaluation in Presence of Biased Noisy Labels %A Riccardo Fogliato %A Alexandra Chouldechova %A Max G’Sell %B Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2020 %E Silvia Chiappa %E Roberto Calandra %F pmlr-v108-fogliato20a %I PMLR %P 2325--2336 %U http://proceedings.mlr.press/v108/fogliato20a.html %V 108 %X Risk assessment tools are widely used around the country to inform decision making within the criminal justice system. Recently, considerable attention has been devoted to the question of whether such tools may suffer from racial bias. In this type of assessment, a fundamental issue is that the training and evaluation of the model is based on a variable (arrest) that may represent a noisy version of an unobserved outcome of more central interest (offense). We propose a sensitivity analysis framework for assessing how assumptions on the noise across groups affect the predictive bias properties of the risk assessment model as a predictor of reoffense. Our experimental results on two real world criminal justice data sets demonstrate how even small biases in the observed labels may call into question the conclusions of an analysis based on the noisy outcome.
APA
Fogliato, R., Chouldechova, A. & G’Sell, M.. (2020). Fairness Evaluation in Presence of Biased Noisy Labels. Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 108:2325-2336 Available from http://proceedings.mlr.press/v108/fogliato20a.html .

Related Material