Safe machine learning model release from Trusted Research Environments: The SACRO-ML package
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2022
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866909714383634432 |
|---|---|
| author | Smith, Jim Preen, Richard J. McCarthy, Andrew Albashir, Maha Crespi-Boixader, Alba Mumtaz, Shahzad Cole, Christian Liley, James Migenda, Jost Rogers, Simon Jones, Yola |
| author_facet | Smith, Jim Preen, Richard J. McCarthy, Andrew Albashir, Maha Crespi-Boixader, Alba Mumtaz, Shahzad Cole, Christian Liley, James Migenda, Jost Rogers, Simon Jones, Yola |
| contents | We present SACRO-ML, an integrated suite of open source Python tools to facilitate the statistical disclosure control (SDC) of machine learning (ML) models trained on confidential data prior to public release. SACRO-ML combines (i) a SafeModel package that extends commonly used ML models to provide ante-hoc SDC by assessing the vulnerability of disclosure posed by the training regime; and (ii) an Attacks package that provides post-hoc SDC by rigorously assessing the empirical disclosure risk of a model through a variety of simulated attacks after training. The SACRO-ML code and documentation are available under an MIT license at https://github.com/AI-SDC/SACRO-ML |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2212_01233 |
| institution | arXiv |
| publishDate | 2022 |
| record_format | arxiv |
| spellingShingle | Safe machine learning model release from Trusted Research Environments: The SACRO-ML package Smith, Jim Preen, Richard J. McCarthy, Andrew Albashir, Maha Crespi-Boixader, Alba Mumtaz, Shahzad Cole, Christian Liley, James Migenda, Jost Rogers, Simon Jones, Yola Machine Learning Cryptography and Security Information Retrieval We present SACRO-ML, an integrated suite of open source Python tools to facilitate the statistical disclosure control (SDC) of machine learning (ML) models trained on confidential data prior to public release. SACRO-ML combines (i) a SafeModel package that extends commonly used ML models to provide ante-hoc SDC by assessing the vulnerability of disclosure posed by the training regime; and (ii) an Attacks package that provides post-hoc SDC by rigorously assessing the empirical disclosure risk of a model through a variety of simulated attacks after training. The SACRO-ML code and documentation are available under an MIT license at https://github.com/AI-SDC/SACRO-ML |
| title | Safe machine learning model release from Trusted Research Environments: The SACRO-ML package |
| topic | Machine Learning Cryptography and Security Information Retrieval |
| url | https://arxiv.org/abs/2212.01233 |