Thesis
Knowledge-refined deep neural networks for real-world applications
- Abstract:
-
Deep neural networks have demonstrated remarkable success in finding patterns in raw data, yet their purely data-driven nature often results in predictions that violate fundamental background knowledge. In this thesis, I address this critical problem by proposing novel neuro-symbolic methods, which integrate background knowledge constraints into neural networks for real-world applications. While background knowledge broadly refers to established logical rules or physical constraints that gove...
Expand abstract
Actions
Access Document
- Files:
-
-
(Preview, Dissemination version, pdf, 17.3MB, Terms of use)
-
Authors
Contributors
+ Lukasiewicz, T
- Institution:
- University of Oxford
- Division:
- MPLS
- Department:
- Computer Science
- Role:
- Supervisor
- ORCID:
- 0000-0002-7644-1668
+ Calinescu, A
- Institution:
- University of Oxford
- Division:
- MPLS
- Department:
- Computer Science
- Role:
- Examiner
+ Minervini, P
- Role:
- Examiner
+ Engineering and Physical Sciences Research Council
More from this funder
- Funder identifier:
- https://ror.org/0439y7842
- Grant:
- EP/T517811/1
- Programme:
- Engineering and Physical Sciences Research Council Scholarship for Doctoral Studies
- DOI:
- Type of award:
- DPhil
- Level of award:
- Doctoral
- Awarding institution:
- University of Oxford
- Language:
-
English
- Keywords:
- Subjects:
- Deposit date:
-
2026-06-02
- ARK identifier:
Terms of use
- Copyright holder:
- Mihaela Cǎtǎlina Stoian
- Copyright date:
- 2025
- Licence:
- CC Attribution (CC BY)
If you are the owner of this record, you can report an update to it here: Report update to this record