Open AccessPosted Content
Data Decisions and Theoretical Implications when Adversarially Learning Fair Representations
Reads0
Chats0
TLDR
An adversarial training procedure is used to remove information about the sensitive attribute from the latent representation learned by a neural network, and the data distribution empirically drives the adversary's notion of fairness.Abstract:
How can we learn a classifier that is "fair" for a protected or sensitive group, when we do not know if the input to the classifier belongs to the protected group? How can we train such a classifier when data on the protected group is difficult to attain? In many settings, finding out the sensitive input attribute can be prohibitively expensive even during model training, and sometimes impossible during model serving. For example, in recommender systems, if we want to predict if a user will click on a given recommendation, we often do not know many attributes of the user, e.g., race or age, and many attributes of the content are hard to determine, e.g., the language or topic. Thus, it is not feasible to use a different classifier calibrated based on knowledge of the sensitive attribute.
Here, we use an adversarial training procedure to remove information about the sensitive attribute from the latent representation learned by a neural network. In particular, we study how the choice of data for the adversarial training effects the resulting fairness properties. We find two interesting results: a small amount of data is needed to train these adversarial models, and the data distribution empirically drives the adversary's notion of fairness.read more
Citations
More filters
Proceedings ArticleDOI
Pushing the Accuracy-Group Robustness Frontier with Introspective Self-play
Jeremiah Zhe Liu,Krishnamurthy Dvijotham,Jihyeon Janel Lee,Quan Yuan,Martin Strobel,Balaji Lakshminarayanan,Deepak Ramachandran +6 more
TL;DR: This paper proposed Introspective Self-play (ISP) to improve the uncertainty estimation of a deep neural network under dataset bias, by adding an auxiliary introspection task requiring a model to predict the bias for each data point in addition to the label.
Posted Content
Impossibility results for fair representations
TL;DR: The authors showed that no representation can guarantee the fairness of classifiers for different tasks trained using it; even the basic goal of achieving label-independent Demographic Parity fairness fails once the marginal data distribution shifts.
Posted Content
The Price of Fair PCA: One Extra Dimension
TL;DR: The notion of Fair PCA is defined and a polynomial-time algorithm for finding a low dimensional representation of the data which is nearly-optimal with respect to this measure is given.
Posted Content
Technical Challenges for Training Fair Neural Networks.
TL;DR: In this article, the authors observe that these large models overfit to fairness objectives, and produce a range of unintended and undesirable consequences, and conduct their experiments on both facial recognition and automated medical diagnosis datasets using state-of-the-art architectures.
Posted Content
AI Fairness via Domain Adaptation.
Neil Joshi,Philippe Burlina +1 more
TL;DR: In this paper, a novel method for ensuring AI fairness with respect to protected or sensitive factors is proposed, which uses domain adaptation via training set enhancement to tackle bias-causing training data imbalance.
References
More filters
Proceedings Article
Adaptive Subgradient Methods for Online Learning and Stochastic Optimization.
TL;DR: Adaptive subgradient methods as discussed by the authors dynamically incorporate knowledge of the geometry of the data observed in earlier iterations to perform more informative gradient-based learning, which allows us to find needles in haystacks in the form of very predictive but rarely seen features.
Journal Article
Adaptive Subgradient Methods for Online Learning and Stochastic Optimization
TL;DR: This work describes and analyze an apparatus for adaptively modifying the proximal function, which significantly simplifies setting a learning rate and results in regret guarantees that are provably as good as the best proximal functions that can be chosen in hindsight.
Book ChapterDOI
Domain-adversarial training of neural networks
Yaroslav Ganin,Evgeniya Ustinova,Hana Ajakan,Pascal Germain,Hugo Larochelle,François Laviolette,Mario Marchand,Victor Lempitsky +7 more
TL;DR: In this article, a new representation learning approach for domain adaptation is proposed, in which data at training and test time come from similar but different distributions, and features that cannot discriminate between the training (source) and test (target) domains are used to promote the emergence of features that are discriminative for the main learning task on the source domain.
Proceedings Article
Equality of opportunity in supervised learning
TL;DR: This work proposes a criterion for discrimination against a specified sensitive attribute in supervised learning, where the goal is to predict some target based on available features and shows how to optimally adjust any learned predictor so as to remove discrimination according to this definition.