Metraipolis › archive › dc_3214b154ddc8
Portable Causal Fairness Across Synthetic Data Generator Families — arXiv:2609.03180
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.03180
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T02:07:49.418Z
- SHA-256 of the archived text
42c54241e63283967ab5fd7615e8bba72b7231fd033acbc69380b4bd0e9add50- Stable URL of this page
https://metraipolis.com/documento/dc_3214b154ddc8
How to cite
Portable Causal Fairness Across Synthetic Data Generator Families — arXiv:2609.03180 — arXiv (en), https://arxiv.org/abs/2609.03180 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_3214b154ddc8 (Metraipolis archive dc_3214b154ddc8, sha256 42c54241e63283967ab5fd7615e8bba72b7231fd033acbc69380b4bd0e9add50)
BibTeX
@misc{dc3214b154ddc8,
title = {Portable Causal Fairness Across Synthetic Data Generator Families — arXiv:2609.03180},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.03180},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_3214b154ddc8; sha256 42c54241e63283967ab5fd7615e8bba72b7231fd033acbc69380b4bd0e9add50},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.03180. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
When a statistical agency or regulator releases synthetic data in place of sensitive records, it chooses the generator that produces the table, and can shape that generator so unfair pathways are absent. DECAF made this concrete on one non-private GAN: three fairness definitions become three sets of edge cuts on the generator's causal graph. Whether the mechanism belongs to DECAF, or to causal factorisation itself, was untested. We port all three definitions to nine generators from three unrelated families (marginals-based, GAN, and diffusion, each with differentially private variants), across three levels of formal privacy guarantee, over 2,520 matched-pair runs on Adult and COMPAS datasets. The mechanism transfers everywhere, and our new causal diffusion backbone yields the fairest release of any family we tested, at fidelity close to the marginals tier. Applying the cut barely moves fidelity, only costs a downstream classifier about $0.07$ to $0.15$ AUC on average, and adding privacy guarantees don't make the data less fair. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-04 in cs.LG). Categorie arXiv: Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2609.03180 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-17T01:47:25.528Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-04?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).