Generalizing Constraint Models in Constraint Acquisition
Dimos Tsouros, Senne Berden, Steven Prestwich Tias Guns · 2024
Read Generalizing Constraint Models in Constraint Acquisition by Dimos Tsouros, Senne Berden, Steven Prestwich Tias Guns (2024) 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-2412-14950.txt.
Description
Dimos Tsouros, Senne Berden, Steven Prestwich, and Tias Guns teach constraint software to learn the rule, not just the instance. Their 2024 paper works on constraint acquisition, systems that learn a problem's constraints from examples so non-experts can model with them. The standard flaw is narrowness: the learner returns constraints for one fixed instance, useless when the numbers change.
Their method, GenCon, learns parameterized constraints instead. A classifier is trained to judge, for any candidate constraint with any parameter values, whether it belongs to the problem. From some classifier types the authors can extract readable rules, producing a specification that generates correct constraints for any new instance on demand. A generate-and-test route covers the other classifiers. Experiments show high accuracy that survives noisy input. The authors frame the advance as the difference between memorizing one puzzle's answer key and learning the puzzle's grammar.
Catalog record
| CREATOR | Dimos Tsouros, Senne Berden, Steven Prestwich Tias Guns |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 52 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2412-14950