Given an algorithmic predictor that is "fair" on some source distribution, will it still be fair on an unknown target distribution that differs from the source within some bound? In this paper, we study the transferability of statistical group fairness for machine learning predictors (i.e., classifiers or regressors) subject to bounded distribution shifts. Such shifts may be introduced by initial training data uncertainties, user adaptation to a deployed predictor, dynamic environments, or the use of pre-trained models in new settings. Herein, we develop a bound that characterizes such transferability, flagging potentially inappropriate deployments of machine learning for socially consequential tasks. We first develop a framework for bounding violations of statistical fairness subject to distribution shift, formulating a generic upper bound for transferred fairness violations as our primary result. We then develop bounds for specific worked examples, focusing on two commonly used fairness definitions (i.e., demographic parity and equalized odds) and two classes of distribution shift (i.e., covariate shift and label shift). Finally, we compare our theoretical bounds to deterministic models of distribution shift and against real-world data, finding that we are able to estimate fairness violation bounds in practice, even when simplifying assumptions are only approximately satisfied.
more »
« less
Robust Fairness under Covariate Shift.
Making predictions that are fair with regard to protected attributes (race, gender, age, etc.) has become an important
requirement for classification algorithms. Existing techniques
derive a fair model from sampled labeled data relying on the
assumption that training and testing data are identically and
independently drawn (iid) from the same distribution. In practice, distribution shift can and does occur between training and
testing datasets as the characteristics of individuals interacting
with the machine learning system change. We investigate fairness under covariate shift, a relaxation of the iid assumption
in which the inputs or covariates change while the conditional
label distribution remains the same. We seek fair decisions under these assumptions on target data with unknown labels. We
propose an approach that obtains the predictor that is robust
to the worst-case testing performance while satisfying target
fairness requirements and matching statistical properties of
the source data. We demonstrate the benefits of our approach
on benchmark prediction tasks.
more »
« less
- Award ID(s):
- 1939743
- NSF-PAR ID:
- 10313981
- Date Published:
- Journal Name:
- Proceedings of the AAAI Conference on Artificial Intelligence
- ISSN:
- 2159-5399
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Spiking neural networks(SNNs) have drawn broad research interests in recent years due to their high energy efficiency and biologically-plausibility. They have proven to be competitive in many machine learning tasks. Similar to all Artificial Neural Network(ANNs) machine learning models, the SNNs rely on the assumption that the training and testing data are drawn from the same distribution. As the environment changes gradually, the input distribution will shift over time, and the performance of SNNs turns out to be brittle. To this end, we propose a unified framework that can adapt nonstationary streaming data by exploiting unlabeled intermediate domain, and fits with the in-hardware SNN learning algorithm Error-modulated STDP. Specifically, we propose a unique self training framework to generate pseudo labels to retrain the model for intermediate and target domains. In addition, we develop an online-normalization method with an auxiliary neuron to normalize the output of the hidden layers. By combining the normalization with self-training, our approach gains average classification improvements over 10% on MNIST, NMINST, and two other datasets.more » « less
-
Spiking neural networks(SNNs) have drawn broad research interests in recent years due to their high energy efficiency and biologically-plausibility. They have proven to be competitive in many machine learning tasks. Similar to all Artificial Neural Network(ANNs) machine learning models, the SNNs rely on the assumption that the training and testing data are drawn from the same distribution. As the environment changes gradually, the input distribution will shift over time, and the performance of SNNs turns out to be brittle. To this end, we propose a unified framework that can adapt non-stationary streaming data by exploiting unlabeled intermediate domain, and fits with the in-hardware SNN learning algorithm Error-modulated STDP. Specifically, we propose a unique self-training framework to generate pseudo labels to retrain the model for intermediate and target domains. In addition, we develop an online-normalization method with an auxiliary neuron to normalize the output of the hidden layers. By combining the normalization with self-training, our approach gains average classification improvements over 10% on MNIST, NMINST, and two other datasets.more » « less
-
As machine learning (ML) algorithms are increasingly used in high-stakes applications, concerns have arisen that they may be biased against certain social groups. Although many approaches have been proposed to make ML models fair, they typically rely on the assumption that data distributions in training and deployment are identical. Unfortunately, this is commonly violated in practice and a model that is fair during training may lead to an unexpected outcome during its deployment. Although the problem of designing robust ML models under dataset shifts has been widely studied, most existing works focus only on the transfer of accuracy. In this paper, we study the transfer of both fairness and accuracy under domain generalization where the data at test time may be sampled from never-before-seen domains. We first develop theoretical bounds on the unfairness and expected loss at deployment, and then derive sufficient conditions under which fairness and accuracy can be perfectly transferred via invariant representation learning. Guided by this, we design a learning algorithm such that fair ML models learned with training data still have high fairness and accuracy when deployment environments change. Experiments on real-world data validate the proposed algorithm.more » « less
-
Recent work in fairness in machine learning has proposed adjusting for fairness by equalizing accuracy metrics across groups and has also studied how datasets affected by historical prejudices may lead to unfair decision policies. We connect these lines of work and study the residual unfairness that arises when a fairness-adjusted predictor is not actually fair on the target population due to systematic censoring of training data by existing biased policies. This scenario is particularly common in the same applications where fairness is a concern. We characterize theoretically the impact of such censoring on standard fairness metrics for binary classifiers and provide criteria for when residual unfairness may or may not appear. We prove that, under certain conditions, fairness-adjusted classifiers will in fact induce residual unfairness that perpetuates the same injustices, against the same groups, that biased the data to begin with, thus showing that even state-of-the-art fair machine learning can have a "bias in, bias out" property. When certain benchmark data is available, we show how sample reweighting can estimate and adjust fairness metrics while accounting for censoring. We use this to study the case of Stop, Question, and Frisk (SQF) and demonstrate that attempting to adjust for fairness perpetuates the same injustices that the policy is infamous for.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
