EUVIMEDEUVIMED entdecken
Crossref

On the Probabilistic Learnability of Compact Neural Network Preimage Bounds

Luca Marzari, Manuele Bicego, Ferdinando Cicalese, Alessandro Farinelli

Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (42) · 35707-35714

DOI: 10.1609/aaai.v40i42.40883 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Luca Marzari und 3 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Although recent provable methods have been developed to compute preimage bounds for neural networks, their scalability is fundamentally limited by the #P-hardness of the problem. In this work, we adopt a novel probabilistic perspective, aiming to deliver solutions with high-confidence guarantees and bounded error. To this end, we investigate the potential of bootstrap-based and randomized approaches that are capable of capturing complex patterns in high-dimensional spaces, including input regions where a given output property holds. In detail, we introduce Random Forest Property Verifier (RF-ProVe), a method that exploits an ensemble of randomized decision trees to generate candidate input regions satisfying a desired output property and refines them through active resampling. Our theoretical derivations offer formal statistical guarantees on region purity and global coverage, providing a practical, scalable solution for computing compact preimage approximations in cases where exact solvers fail to scale.

Zitieren

APA

Marzari, L., Bicego, M., Cicalese, F., & Farinelli, A. (2026). On the Probabilistic Learnability of Compact Neural Network Preimage Bounds. Proceedings of the AAAI Conference on Artificial Intelligence, 40(42), 35707-35714. https://doi.org/10.1609/aaai.v40i42.40883

Vancouver

Marzari L, Bicego M, Cicalese F, Farinelli A. On the Probabilistic Learnability of Compact Neural Network Preimage Bounds. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(42):35707-35714. doi:10.1609/aaai.v40i42.40883

BibTeX
@article{Marzari2026,
  author = {Marzari, Luca and Bicego, Manuele and Cicalese, Ferdinando and Farinelli, Alessandro},
  title = {On the Probabilistic Learnability of Compact Neural Network Preimage Bounds},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {42},
  pages = {35707--35714},
  doi = {10.1609/aaai.v40i42.40883}
}
Kostenlos starten →