Aisdc

Latest version: v1.1.2

Safety actively analyzes 613705 Python packages for vulnerabilities to keep your Python projects secure.

Scan your dependencies

Page 1 of 2

1.1.2

Changes:
* Fix a bug related to the `rules.json` path when running from package ([247](https://github.com/AI-SDC/AI-SDC/pull/247))
* Update user stories ([247](https://github.com/AI-SDC/AI-SDC/pull/247))

1.1.1

Changes:
* Update notebook example paths ([237](https://github.com/AI-SDC/AI-SDC/pull/237))
* Fix AdaBoostClassifier structural attack ([242](https://github.com/AI-SDC/AI-SDC/pull/242))
* Move experiments module and configs to separate repository ([229](https://github.com/AI-SDC/AI-SDC/pull/229))

1.1.0

Changes:
* Add automatic formatting of docstrings ([210](https://github.com/AI-SDC/AI-SDC/pull/210))
* Update user stories ([217](https://github.com/AI-SDC/AI-SDC/pull/217))
* Add module to run experiments with attacks and gather data ([224](https://github.com/AI-SDC/AI-SDC/pull/224))
* Fix bug in report.py: error removing a file that does not exist ([227](https://github.com/AI-SDC/AI-SDC/pull/227))
* Add structural attack for traditional and other risk measures ([232](https://github.com/AI-SDC/AI-SDC/pull/232))
* Fix package installation for Python 3.8, 3.9, 3.10 ([234](https://github.com/AI-SDC/AI-SDC/pull/234))

1.0.6

Changes:
* Update package dependencies ([187](https://github.com/AI-SDC/AI-SDC/pull/187))
* Fix bug when `n_dummy_reps=0` in worst case attack ([191](https://github.com/AI-SDC/AI-SDC/pull/191))
* Add ability to save target model and data to `target.json` ([171](https://github.com/AI-SDC/AI-SDC/pull/171), [#175](https://github.com/AI-SDC/AI-SDC/pull/175), [#176](https://github.com/AI-SDC/AI-SDC/pull/176), [#177](https://github.com/AI-SDC/AI-SDC/pull/177))
* Add safemodel SDC results to `target.json` and `attack_results.json` ([180](https://github.com/AI-SDC/AI-SDC/pull/180))
* Add generalisation error to `target.json` ([183](https://github.com/AI-SDC/AI-SDC/pull/183))
* Refactor attack argument handling ([174](https://github.com/AI-SDC/AI-SDC/pull/174))
* Append attack outputs to a single results file ([173](https://github.com/AI-SDC/AI-SDC/pull/173))
* Attack outputs written to specified folder ([208](https://github.com/AI-SDC/AI-SDC/pull/208))
* Add ability to run membership inference attacks from the command line using config and target files ([182](https://github.com/AI-SDC/AI-SDC/pull/182))
* Add ability to run attribute inference attacks from the command line using config and target files ([188](https://github.com/AI-SDC/AI-SDC/pull/188))
* Add ability to run multiple attacks from a config file ([200](https://github.com/AI-SDC/AI-SDC/pull/200))
* Add user story examples ([194](https://github.com/AI-SDC/AI-SDC/pull/194))
* Improve attack formatter summary generation ([179](https://github.com/AI-SDC/AI-SDC/pull/179))
* Attack formatter moves files generated for release into subfolders ([197](https://github.com/AI-SDC/AI-SDC/pull/197))
* Fix a minor bug in the attack formatter ([204](https://github.com/AI-SDC/AI-SDC/pull/204))
* Improve tests ([196](https://github.com/AI-SDC/AI-SDC/pull/196), [#199](https://github.com/AI-SDC/AI-SDC/pull/199))

1.0.5

Changes:
* Fix a bug calculating the number of data samples in the `Data` class ([105](https://github.com/AI-SDC/AI-SDC/pull/105))
* Add a fail-fast mechanism for the worst case attack that enables the number of attack repetitions to terminate early based on a given metric and comparison operator ([105](https://github.com/AI-SDC/AI-SDC/pull/105))
* Change the logging message when attack repetitions are run to 1-10 instead of 0-9 ([105](https://github.com/AI-SDC/AI-SDC/pull/105))
* Add the ability to specify the number of worst case attack dummy repetitions on the command line ([105](https://github.com/AI-SDC/AI-SDC/pull/105))
* Add LIRA fail-fast mechanism ([118](https://github.com/AI-SDC/AI-SDC/pull/118))
* Add the ability to load LIRA attack parameters from a config file ([118](https://github.com/AI-SDC/AI-SDC/pull/118))
* Add the ability to load worst case attack parameters from a config file ([119](https://github.com/AI-SDC/AI-SDC/pull/119))
* Standardise the MIA attack output ([120](https://github.com/AI-SDC/AI-SDC/pull/120))
* Prohibit the use of white space in report file names ([154](https://github.com/AI-SDC/AI-SDC/pull/154))
* Improve the safemodel request release test ([160](https://github.com/AI-SDC/AI-SDC/pull/160))
* Refactor LIRA attack tests ([151](https://github.com/AI-SDC/AI-SDC/pull/151))
* Fix setting the number of LIRA shadow models from a config file ([165](https://github.com/AI-SDC/AI-SDC/pull/165))
* Fix OS system calls relying on calling "python" ([162](https://github.com/AI-SDC/AI-SDC/pull/162))
* Fix invalid command line argument in worst case attack example ([164](https://github.com/AI-SDC/AI-SDC/pull/164))
* Add current output JSON format documentation ([168](https://github.com/AI-SDC/AI-SDC/pull/168))
* Add current attack config format documentation ([168](https://github.com/AI-SDC/AI-SDC/pull/168))

1.0.4

Changes:
* Fixed SafeRandomForestClassifier "base estimator changed" error ([143](https://github.com/AI-SDC/AI-SDC/pull/143))

Page 1 of 2

© 2024 Safety CLI Cybersecurity Inc. All Rights Reserved.