We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

stat

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: A Unified Approach to Fair Online Learning via Blackwell Approachability

Authors: Evgenii Chzhen (LMO, CELESTE), Christophe Giraud (LMO, CELESTE), Gilles Stoltz (LMO, CELESTE)
Abstract: We provide a setting and a general approach to fair online learning with stochastic sensitive and non-sensitive contexts. The setting is a repeated game between the Player and Nature, where at each stage both pick actions based on the contexts. Inspired by the notion of unawareness, we assume that the Player can only access the non-sensitive context before making a decision, while we discuss both cases of Nature accessing the sensitive contexts and Nature unaware of the sensitive contexts. Adapting Blackwell's approachability theory to handle the case of an unknown contexts' distribution, we provide a general necessary and sufficient condition for learning objectives to be compatible with some fairness constraints. This condition is instantiated on (group-wise) no-regret and (group-wise) calibration objectives, and on demographic parity as an additional constraint. When the objective is not compatible with the constraint, the provided framework permits to characterise the optimal trade-off between the two.
Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Machine Learning (stat.ML)
Journal reference: 35th Conference on Neural Information Processing Systems (NeurIPS 2021), Dec 2021, Virtual conference
Cite as: arXiv:2106.12242 [cs.LG]
  (or arXiv:2106.12242v2 [cs.LG] for this version)

Submission history

From: Gilles Stoltz [view email]
[v1] Wed, 23 Jun 2021 09:00:12 GMT (75kb)
[v2] Sun, 7 Nov 2021 21:14:04 GMT (69kb)

Link back to: arXiv, form interface, contact.