A General Framework for Property-Driven Machine Learning
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912445741662208 |
|---|---|
| author | Flinkow, Thomas Casadio, Marco Kessler, Colin Monahan, Rosemary Komendantskaya, Ekaterina |
| author_facet | Flinkow, Thomas Casadio, Marco Kessler, Colin Monahan, Rosemary Komendantskaya, Ekaterina |
| contents | Neural networks have been shown to frequently fail to learn critical safety and correctness properties purely from data, highlighting the need for training methods that directly integrate logical specifications. While adversarial training can be used to improve robustness to small perturbations within $ε$-cubes, domains other than computer vision -- such as control systems and natural language processing -- may require more flexible input region specifications via generalised hyper-rectangles. Differentiable logics offer a way to encode arbitrary logical constraints as additional loss terms that guide the learning process towards satisfying these constraints. In this paper, we investigate how these two complementary approaches can be unified within a single framework for property-driven machine learning, as a step toward effective formal verification of neural networks. We show that well-known properties from the literature are subcases of this general approach, and we demonstrate its practical effectiveness on a case study involving a neural network controller for a drone system. Our framework is made publicly available at https://github.com/tflinkow/property-driven-ml. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_00466 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | A General Framework for Property-Driven Machine Learning Flinkow, Thomas Casadio, Marco Kessler, Colin Monahan, Rosemary Komendantskaya, Ekaterina Machine Learning Logic in Computer Science Neural networks have been shown to frequently fail to learn critical safety and correctness properties purely from data, highlighting the need for training methods that directly integrate logical specifications. While adversarial training can be used to improve robustness to small perturbations within $ε$-cubes, domains other than computer vision -- such as control systems and natural language processing -- may require more flexible input region specifications via generalised hyper-rectangles. Differentiable logics offer a way to encode arbitrary logical constraints as additional loss terms that guide the learning process towards satisfying these constraints. In this paper, we investigate how these two complementary approaches can be unified within a single framework for property-driven machine learning, as a step toward effective formal verification of neural networks. We show that well-known properties from the literature are subcases of this general approach, and we demonstrate its practical effectiveness on a case study involving a neural network controller for a drone system. Our framework is made publicly available at https://github.com/tflinkow/property-driven-ml. |
| title | A General Framework for Property-Driven Machine Learning |
| topic | Machine Learning Logic in Computer Science |
| url | https://arxiv.org/abs/2505.00466 |