Learning Formal Specifications from Membership and Preference Queries
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2023
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866909620416544768 |
|---|---|
| author | Shah, Ameesh Vazquez-Chanlatte, Marcell Junges, Sebastian Seshia, Sanjit A. |
| author_facet | Shah, Ameesh Vazquez-Chanlatte, Marcell Junges, Sebastian Seshia, Sanjit A. |
| contents | Active learning is a well-studied approach to learning formal specifications, such as automata. In this work, we extend active specification learning by proposing a novel framework that strategically requests a combination of membership labels and pair-wise preferences, a popular alternative to membership labels. The combination of pair-wise preferences and membership labels allows for a more flexible approach to active specification learning, which previously relied on membership labels only. We instantiate our framework in two different domains, demonstrating the generality of our approach. Our results suggest that learning from both modalities allows us to robustly and conveniently identify specifications via membership and preferences. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2307_10434 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Learning Formal Specifications from Membership and Preference Queries Shah, Ameesh Vazquez-Chanlatte, Marcell Junges, Sebastian Seshia, Sanjit A. Formal Languages and Automata Theory Artificial Intelligence Machine Learning Active learning is a well-studied approach to learning formal specifications, such as automata. In this work, we extend active specification learning by proposing a novel framework that strategically requests a combination of membership labels and pair-wise preferences, a popular alternative to membership labels. The combination of pair-wise preferences and membership labels allows for a more flexible approach to active specification learning, which previously relied on membership labels only. We instantiate our framework in two different domains, demonstrating the generality of our approach. Our results suggest that learning from both modalities allows us to robustly and conveniently identify specifications via membership and preferences. |
| title | Learning Formal Specifications from Membership and Preference Queries |
| topic | Formal Languages and Automata Theory Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2307.10434 |