Exploring the Unfairness of DP-SGD Across Settings
Research output: Contribution to conference › Paper › Research
Standard
Exploring the Unfairness of DP-SGD Across Settings. / Noe, Frederik ; Herskind , Rasmus ; Søgaard, Anders.
2022. Paper presented at Third AAAI Workshop on Privacy-Preserving Artificial Intelligence (PPAI-22), VIRTUAL.Research output: Contribution to conference › Paper › Research
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - CONF
T1 - Exploring the Unfairness of DP-SGD Across Settings
AU - Noe, Frederik
AU - Herskind , Rasmus
AU - Søgaard, Anders
PY - 2022
Y1 - 2022
N2 - End users and regulators require private and fair artificial intelligence models, but previous work suggests these objectives may be at odds. We use the CivilComments to evaluate the impact of applying the {\em de facto} standard approach to privacy, DP-SGD, across several fairness metrics. We evaluate three implementations of DP-SGD: for dimensionality reduction (PCA), linear classification (logistic regression), and robust deep learning (Group-DRO). We establish a negative, logarithmic correlation between privacy and fairness in the case of linear classification and robust deep learning. DP-SGD had no significant impact on fairness for PCA, but upon inspection, also did not seem to lead to private representations.
AB - End users and regulators require private and fair artificial intelligence models, but previous work suggests these objectives may be at odds. We use the CivilComments to evaluate the impact of applying the {\em de facto} standard approach to privacy, DP-SGD, across several fairness metrics. We evaluate three implementations of DP-SGD: for dimensionality reduction (PCA), linear classification (logistic regression), and robust deep learning (Group-DRO). We establish a negative, logarithmic correlation between privacy and fairness in the case of linear classification and robust deep learning. DP-SGD had no significant impact on fairness for PCA, but upon inspection, also did not seem to lead to private representations.
M3 - Paper
T2 - Third AAAI Workshop on Privacy-Preserving Artificial Intelligence (PPAI-22)
Y2 - 28 February 2022
ER -
ID: 341484877