Exact characterization of ε \varepsilon -Safe Decision Regions for exponential family distributions and Multi Cost SVM approximation
Alberto Carlevaro [email protected], Teodoro Alamo [email protected], Fabrizio Dabbene [email protected], Maurizio Mongelli [email protected] · 2025
Read Exact characterization of ε \varepsilon -Safe Decision Regions for exponential family distributions and Multi Cost SVM approximation by Alberto Carlevaro [email protected], Teodoro Alamo [email protected], Fabrizio Dabbene [email protected], Maurizio Mongelli [email protected] (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-17731.txt.
Description
Alberto Carlevaro, Teodoro Alamo, Fabrizio Dabbene, and Maurizio Mongelli pin down safe decision zones for classifiers, where a safe class call is backed by a chance bound. Trust in machine learning rests on a clear line between safe and unsafe calls. Their work came out in twenty twenty-five.
They define such a zone in formal terms, set by the chance of the target class and the faith put in the call. When data come from a broad family of distributions, the zone's shape is fixed by math and can be tuned by design. Since real data may break that family, they add a support vector method that mimics the zone and handles lopsided data. Tests and shared code back the claims.
Catalog record
| CREATOR | Alberto Carlevaro [email protected], Teodoro Alamo [email protected], Fabrizio Dabbene [email protected], Maurizio Mongelli [email protected] |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 58 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-17731