Learning Fair Representations

By Richard S. Zemel, Ledell Yu Wu, Kevin Swersky, Toniann Pitassi, Cynthia Dwork. 2013

2,047 citations recorded by the source corpus.

Subfields: Safety, Robustness & Interpretability

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Safety, Robustness & Interpretability

Citations in the corpus

Reducing Sentiment Bias in Language Models via Counterfactual Evaluation (2019, background): “We aim towards counterfactual fairness by debiasing the latent representation of inputs in the language models, contributing to a family of methods to learn fair representations (Beutel et al., 2017; Zemel et al., 2013; Creager et al., 2019; Edwards and Storkey, 2016; Louizos et al., 2016) and enfor”
Shortcut learning in deep neural networks (2020, background): “according to a certain deﬁnition [132]. Individual fairness aims at treating similar individuals similarly while group fairness aims at treating subgroups no different than the rest of the population [133,134]. Fairness is closely linked to generalisation and causality [135]. Sensitive group members”
FairFace: Face Attribute Dataset for Balanced Race, Gender, and Age (2019, background): “Researchers in AI and machine learning have increasingly paid attention to algorithmic fairness and dataset and model biases [71, 11, 76, 72].”
Semantics derived automatically from language corpora contain human-like biases (2016, background): “A recent line of work on fairness in machine learning tries to minimize or avoid such biases (Dwork et al., 2012; Feldman et al., 2015; Zemel et al., 2013; Barocas and Selbst, 2014).”
Generative Moment Matching Networks (2015, background): “An interesting application of MMD that is not directly re-lated to generative modelling comes from recent work on learning fair representations (Zemel et al., 2013).”
Concrete Problems in AI Safety (2016, background): “• Fairness: How can we make sure ML systems don’t discriminate? [3, 163, 6, 45, 117, 164]”
Does Object Recognition Work for Everyone? (2019, background): “For instance, statistical parity [5, 6, 14, 23–25, 30, 49] states that in binary-classiﬁcation settings, members of different groups should have the same chance of receiving a positive prediction.”
Towards fairer datasets: filtering and balancing the distribution of the people subtree in the ImageNet hierarchy (2019, background): “Algorithmic solutions have ranged from removing undesired bias by preprocessing the data [39, 59], striking a tradeoff between performance and fairness by posing additional regularization during training or inference [34, 40, 52, 80, 81, 83], or designing application-specific interventions (such as ”
Counterfactual Fairness in Text Classification through Robustness (2018, methodology): “d on debiasing methods to address these gaps. Many methods have been proposed to address group fairness issues, such as re-calibrating score functions [5], adversarially learning fair representations [1, 14, 22], data rebalancing [2], and data augmentation using swaps of gender terms [18]. Less sign”
Bias in Bios: A Case Study of Semantic Representation Bias in a High-Stakes Setting (2019, background): “ontext of structured data, there have been extensive discussions about proxy behavior that may occur when sensitive attributes are not explicitly available but can be determined from other attributes [4, 33, 43]. Related discussions have focused on the phenomenon of differential subgroup validity [3”
Certifying and Removing Disparate Impact (2014, methodology): “[25], we will use the same data sets and the same decisions about what constitutes a sensitive attribute as they do.”
Inherent Trade-Offs in the Fair Determination of Risk Scores (2016, background): “look to learn fair intermediate representations of data while preserving information needed for classification [29].”
Explainable Artificial Intelligence (XAI): Concepts, Taxonomies, Opportunities and Challenges toward Responsible AI (2019, neutral): “Another example is [396] , which hinges on transforming the input data in order to ﬁnd a good representation that obfuscates information about membership in sensitive features.”
Counterfactual Fairness (2017, background): “These include fairness through unawareness [12], individual fairness [10, 16, 24, 38], demographic parity/disparate impact [36], and equality of opportunity [14, 37].”
Beyond Individual and Group Fairness (2020, background): “This similarity is often defined according to an underlying metric over user features [Zemel et al., 2013, Dwork et al., 2012, Joseph ∗Google Research and Rutgers University. pranjalawasthi@google.com †Google Research. corinna@google.com ‡Tel Aviv University and Google Research.…”
NoPeek: Information leakage reduction to share activations in distributed deep learning (2020, background): “A similar nonadversarial approach that still uses a dependency measure of maximum mean discrepancy (MMD) for learning a variational autoencoder is in [37, 38].”
Learning Adversarially Fair and Transferable Representations (2018, background): “Zemel et al. (2013) encouraged elements of both group and individual fairness via a regularized objective.”
Contamination Attacks and Mitigation in Multi-Party Machine Learning (2018, background): “s on which the model is conﬁgured to predict the attacker-chosen label should appear naturally in the test set of the victim parties. Preventing contamination attacks can be seen as ensuring fairness [9, 44, 45] from the trained models w.r.t. the contaminated attributes. This line of work assumes th”