Metraipolis › archive › dc_323b07c32bb1
On the Reliability of Generative Augmentation: A Wasserstein-Based Theoretical and Empirical Study — arXiv:2609.01410
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.01410
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T06:52:39.935Z
- SHA-256 of the archived text
6ab0bb3d7ad8521f1868e0bb2d5aa1f481dd160e0fb83a18a6a1c9d86097dbcb- Stable URL of this page
https://metraipolis.com/documento/dc_323b07c32bb1
How to cite
On the Reliability of Generative Augmentation: A Wasserstein-Based Theoretical and Empirical Study — arXiv:2609.01410 — arXiv (en), https://arxiv.org/abs/2609.01410 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_323b07c32bb1 (Metraipolis archive dc_323b07c32bb1, sha256 6ab0bb3d7ad8521f1868e0bb2d5aa1f481dd160e0fb83a18a6a1c9d86097dbcb)
BibTeX
@misc{dc323b07c32bb1,
title = {On the Reliability of Generative Augmentation: A Wasserstein-Based Theoretical and Empirical Study — arXiv:2609.01410},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.01410},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_323b07c32bb1; sha256 6ab0bb3d7ad8521f1868e0bb2d5aa1f481dd160e0fb83a18a6a1c9d86097dbcb},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.01410. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Generative data augmentation is widely used to mitigate class imbalance, yet its theoretical effect on downstream generalization remains poorly understood. In this work, we develop a statistical framework for conditional generative augmentation and analyze its impact on classification risk. We formalize augmentation as a distribution-mixing process and show that the resulting risk distortion is controlled by both the augmentation strength and the class-conditional Wasserstein discrepancy between real and generated distributions. We further derive a capacity-dependent generalization bound based on Rademacher complexity, revealing an explicit trade-off between hypothesis complexity, augmentation intensity, and generative fidelity. Empirically, we evaluate the framework on binary and multiclass imbalanced classification tasks using Conditional GAN and Conditional WGAN-GP augmentation. Across datasets, CWGAN-GP consistently achieves lower Wasserstein discrepancies than CGAN, indicating improved distributional fidelity. However, improved fidelity does not necessarily translate into superior classification performance, with classical oversampling methods often remaining competitive. These findings support the central theoretical prediction that augmentation reliability is governed by distributional approximation error rather than predictive performance alone. Overall, this work establishes generative augmentation as a distributional perturbation process whose reliability can be quantified through Wasserstein-based measures and supported by finite-sample generalization guarantees. The proposed framework provides a principled foundation for evaluating synthetic data quality beyond classification accuracy alone. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-02 in cs.LG). Categorie arXiv: Machine Learning (stat.ML) ; Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2609.01410 — 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 2026-10-17T06:41:55.928Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).