lp_nn_robustness_verification - v0.8.0

Feature

• CITATION: Introduce reference to PySCIPOpt and data (4c1d4bb)
• timing_evaluation notebook: Introduce timing evaluation in jupyter notebook (a84e2cf)

Fix

• pre_processing: Replace zi by thetai in the calculation of xi_i (cc1d9c8)
• Adapt all calls of ZeMASamples to most recent version v0.7.0 of zemaemcannotated (6807f6b)

Documentation

• timing_evaluation notebook: Introduce timing evaluation jupyter notebook (56e7bcb)

See all commits in this version

- Jupyter Notebook
Published by BjoernLudwigPTB over 3 years ago

lp_nn_robustness_verification - v0.7.0

Feature

• pre_processing: Utilize timing module in pre-processing (c7926cb)
• timing: Introduce module to time and store progress (dddd24d)

Documentation

• timing: Introduce timing module into docs (5670c98)

See all commits in this version

- Jupyter Notebook
Published by BjoernLudwigPTB over 3 years ago

lp_nn_robustness_verification - v0.6.1

Fix

• linear_program: Remove any clutter from linear_program after finding same results in all cases (e4d1f2f)
• INSTALL.md: Fix a bunch of formatting and information issues (8a1522b)

Documentation

• examples: Introduce all examples code into docs (5d4ddd3)

See all commits in this version

- Jupyter Notebook
Published by BjoernLudwigPTB over 3 years ago

lp_nn_robustness_verification - v0.6.0

Feature

• CITATION.cff: Introduce DOI (942bea6)

Fix

• README: Finalize README with mandatory sections (0c6d01b)

Documentation

• INSTALL: Finalize installation instructions (cf66be8)
• README: Introduce DOI and GitHub release badge (52f2477)

See all commits in this version

- Jupyter Notebook
Published by BjoernLudwigPTB over 3 years ago

lp_nn_robustness_verification - v0.5.0

Feature

• activation_functions: Introduce Identity activation function (9dac464)
• linear_program: Replace auxiliary variable by direct objective and thus fix residing issue (27021f4)
• linear_program: Introduce generic base class and two classes for original and adapted problem (a17389e)
• data_types: Introduce IndexAndSeed data_type for streamlined implementation of example (03c5123)
• examples: Introduce task id parameter to parallelize and streamline implementation (e82582b)
• examples: Introduce script to find valid combinations of seeds, samples, sizes and depths (954d204)
• data_types: Capture the index of the valid samples from the ZeMA dataset (c50a8f6)

Documentation

• linear_inclusion notebook: Update with most recent version of zemaemcannotated (10e9b69)

See all commits in this version

- Jupyter Notebook
Published by BjoernLudwigPTB over 3 years ago