mlsanityData to 7 Oct 2026

Learning to Ignore: Fair and Task Independent Representations

arXiv 2101.04047 ↗First posted 11 Jan 2021cs.LG cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Training fair machine learning models, aiming for their interpretability and solving the problem of domain shift has gained a lot of interest in the last years. There is a vast amount of work addressing these topics, mostly in separation. In this work we show that they can be seen as a common framework of learning invariant representations. The representations should allow to predict the target while at the same time being invariant to sensitive attributes which split the dataset into subgroups. Our approach is based on the simple observation that it is impossible for any learning algorithm to differentiate samples if they have the same feature representation. This is formulated as an additional loss (regularizer) enforcing a common feature representation across subgroups. We apply it to learn fair models and interpret the influence of the sensitive attribute. Furthermore it can be used for domain adaptation, transferring knowledge and learning effectively from very few examples. In all applications it is essential not only to learn to predict the target, but also to learn what to ignore.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 19 of its 31 references that are in the map. First posted on arXiv 11 Jan 2021.

Builds on

19 of 31 references.

References · 19

  1. 2015Domain-Adversarial Training of Neural Networkscited by 3,023
  2. 2016Equality of Opportunity in Supervised Learningcited by 2,000
  3. 2017Causal Effect Inference with Deep Latent-Variable Modelscited by 408
  4. 2016Fairness Beyond Disparate Treatment & Disparate Impact: Learning Classification without Disparate Mistreatmentcited by 368
  5. 2017Unified Deep Supervised Domain Adaptation and Generalizationcited by 322
  6. 2015Domain Generalization for Object Recognition with Multi-task Autoencoderscited by 253
  7. 2020Large image datasets: A pyrrhic win for computer vision?cited by 237
  8. 2015The Variational Fair Autoencodercited by 223
  9. 2017Data Decisions and Theoretical Implications when Adversarially Learning Fair Representationscited by 165
  10. 2017Controllable Invariance through Adversarial Feature Learningcited by 127
  11. 2017Domain Adaptation by Using Causal Inference to Predict Invariant Conditional Distributionscited by 125
  12. 2018Invariance, Causality and Robustnesscited by 123
7 more
  1. 2019Neural Network Attributions: A Causal Perspectivecited by 77
  2. 2017Unsupervised Domain Adaptation with Similarity Learningcited by 73
  3. 2018Boosting Domain Adaptation by Discovering Latent Domainscited by 46
  4. 2019Transfer of Machine Learning Fairness across Domainscited by 36
  5. 2019Controlling for Biasing Signals in Images for Prognostic Models: Survival Predictions for Lung Cancer with Deep Learningcited by 5
  6. 2019Fairness Violations and Mitigation under Covariate Shiftcited by 5
  7. 2016Counterfactual Prediction with Deep Instrumental Variables Networkscited by 4

Built on by

Not cited yet.