Tags give the ability to mark specific points in history as being important
  • v0.11 protected
    Release v0.11


    • #653 Added new secml.ml.model_zoo package, which provides a zoo of pre-trained SecML models. The list of available models will be greatly expanded in the future. See https://secml.gitlab.io/secml.ml.model_zoo.html for more details.
    • #629 Greatly improved the performance of the grad_f_x method for CClassifier and CPreProcess classes, especially when nested via preprocess attribute.
    • #613 Python version 3.5, 3.6, or 3.7 is now required.

    Requirements (2 changes)

    • #633 The following dependencies are now required: numpy >= 1.17, scipy >= 1.3.1, scikit-learn >= 0.21 matplotlib = 3.
    • #622 Removed dependency on six library.

    Added (5 changes)

    • #539 Added new core interface to get and set the state of an object instance: set_state, get_state, save_state, load_state. The state of an object is a simple human-readable Python dictionary object which stores the data necessary to restore an instance to a specific status. Please not that to guarantee the exact match between the original object instance and the restored one, the standard save/load interface should be used.
    • #647 Added new function core.attr_utils.get_protected which returns a protected attribute from a class (if exists).
    • #629 CClassifier and CPreProcess classes now provide a gradient method, which computes the gradient by doing a forward and a backward pass on the classifier or preprocessor function chain, accepting an optional pre-multiplier w.
    • #539 Added new accessible attributes to multiple classes: CNormalizerMinMax .m .q; CReducerLDA .lda; CClassifierKNN .tr; CClassifierRidge .tr; CClassifierSGD .tr; CClassifierPyTorch .trained.
    • #640 Added random_state parameter to CClassifierDecisionTree.

    Improved (6 changes)

    • #631 Data objects are now stored using protocol 4 by pickle_utils.save. This protocol adds support for very large objects, pickling more kinds of objects, and some data format optimizations.
    • #639 Objective function parameter (objective_function) in CAttackEvasionCleverhans is now correctly populated for the following attacks: CarliniWagnerL2, FastGradientMethod, ProjectedGradientDescent, LBFGS, MomentumIterativeMethod, MadryEtAl, BasicIterativeMethod.
    • #638 The sequence of modifications to the attack point (x_seq parameter) is now correctly populated in CAttackEvasionCleverhans.
    • #595 A pre-trained classifier can now be passed to CClassifierRejectThreshold to avoid running fit twice.
    • #627 Slight improvement of CKernel.gradient() method performance by removing unnecessary calls.
    • #630 Sparse data can now be used in CKernelHistIntersect.

    Changed (2 changes)

    • #616 Renamed CModelCleverhans to _CModelCleverhans as this class is not supposed to be explicitly used.
    • #111 Default value of the parameter tol changed from -inf to None in CClassifierSGD. This change should not alter the classifier behavior when using the default parameters.

    Fixed (8 changes)

    • #611 Fixed CDataloaderMNIST crashing depending on the desired number of samples and digits to load.
    • #652 Number of gradient computations returned by CAttackEvasionCleverhans.grad_eval is now accurate.
    • #650 Fixed CAttackEvasionCleverhans.f_eval wrongly returns the number of gradient evaluations.
    • #637 Fixed checks on y_taget in CAttackEvasionCleverhans which compared the 0 label to untargeted case (y_true = None).
    • #648 Function core.attr_utils.is_public now correctly return False for properties.
    • #649 Fixed wrong use of core.attr_utils.is_public in CCreator and CDatasetHeader.
    • #655 Fixed CClassifierRejectThreshold.n_classes not taking into account the rejected class (label -1).
    • #636 Fixed a TypeError raised by CFigure.clabel() when using matplotlib 3.

    Removed & Deprecated (4 changes)

    • #628 Method is_linear of CClassifier and CNormalizer subclasses is now deprecated.
    • #641 Parameter random_seed of CClassifierLogistic is now deprecated. Use random_state instead.
    • #603 Removed deprecated class CNormalizerMeanSTD.
    • #603 Removed deprecated parameter batch_size from CKernel and subclasses.

    Documentation (4 changes)

    • #625 Reorganized notebooks tutorials into different categories: Machine Learning, Adversarial Machine Learning, and Explainable Machine Learning.
    • #615 Added a tutorial notebook on the use of Cleverhans library wrapper.
    • #607 Settings module secml.settings is now correctly displayed in the docs.
    • #626 Added missing reference to CPlotMetric class in docs.
  • v0.11-rc1 protected
    ae8c9bce · Release v0.11-rc1 ·
  • v0.10 protected
    f99cf9ba · Release v0.10 ·
    Release v0.10


    • #535 Added new package secml.explanation, which provides different methods for explaining machine learning models. See documentation and examples for more information.
    • #584 [beta] Added CModelCleverhans and CAttackEvasionCleverhans to support adversarial attacks from CleverHans, a Python library to benchmark vulnerability of machine learning systems to adversarial examples.

    Requirements (1 change)

    • #580 PyTorch version 1.3 is now supported.

    Added (4 changes)

    • #565 Added new abstract interface CClassifierDNN from which new classes implementing Deep Neural Networks can inherit.
    • #555 Added CNormalizerDNN, which allows using a CClassifierDNN as a preprocessor.
    • #593 Added CDataLoaderTorchDataset, which allows converting a torchvision dataset into a CDataset.
    • #598 Added gradient method for CKernelHistIntersection.

    Improved (6 changes)

    • #562 Extended support of CClassifierPyTorch to nested PyTorch modules.
    • #594 CClassifierPyTorch.load_model() is now able to also load models trained with PyTorch (without using our wrapper). New parameter classes added to the method to match classes to indexes in the loaded model.
    • #579 Left side single row/column broadcast is now supported for sparse vs sparse CArray operations.
    • #582 Improved performance of CNormalizerMeanStd when multiple channels are defined.
    • #576 Vastly improved the performance of kernels by removing loops over samples in many classes and refactoring main routines.
    • #562 Improved grad_f_x computation at a specific layer in CClassifierPyTorch.

    Changed (4 changes)

    • #578 CClassifierPyTorch now inherits from CClassifierDNN. The following changed accordingly: parameter torch_model renamed to model; property layer_shapes is now defined; method save_checkpoint removed.
    • #562 Parameter layer of CClassifierPyTorch.get_layer_output() is now renamed layer_names as a list of layers names is supported (a single layer name is still supported as input). A dictionary is returned if multiple layers are requested. See the documentation for more information.
    • #533 Double initialization in CAttackEvasionPGDLS will now be executed regardless of the classifier type (linear or nonlinear) if the double_init parameter of .run() method is set to True.
    • #591 It is now not required to call the fit method of CNormalizerMeanSTD if fixed mean/std values are used.

    Fixed (4 changes)

    • #561 Fixed CConstraintBox not always applied correctly for float data.
    • #577 Fixed CClassifierPyTorch.decision_function applying preprocess twice.
    • #581 Fixed gradient computation of CKernelChebyshevDistance.
    • #599 Kernels using distances are now based on negative distances (to correctly represent similarity measures). Affected classes are: CKernelChebyshevDistance, CKernelEuclidean.

    Removed & Deprecated (5 changes)

    • #561 Removed parameter precision from CConstraint.is_violated().
    • #575 Parameter batch_size of CKernel is now deprecated.
    • #597 Removed unused parameter gamma from CKernelChebyshevDistance.
    • #596 Removed CKernelHamming.
    • #602 Renamed CNormalizerMeanSTD to CNormalizerMeanStd. The old class has been deprecated and will be removed in a future vearsion.

    Documentation (5 changes)

    • #538 Added a notebook tutorial on the use of Explainable ML methods provided by the secml.explanation package.
    • #573 Improved visualization of attack results in 07-ImageNet tutorial.
    • #610 Fixed spacing between parameter and parameter type in the docs.
    • #605 Fixed documentation of classes requiring extra components not being displayed.
    • #608 Added acknowledgments to README.
  • v0.10-rc1 protected
    9e66eca1 · Release v0.10-rc1 ·
  • v0.9 protected
    Release v0.9


    • #536 Added CClassifierPytorch to support Neural Networks (NNs) through PyTorch deep learning platform.

    Improved (1 change)

    • #556 CFigure.imshow now supports PIL images as input.

    Changed (1 change)

    • #532 Method CPreProcess.revert() is now renamed .inverse_transform().

    Fixed (1 change)

    • #554 Fixed CClassifierSkLearn.predict() not working when using pretrained sklearn models.

    Documentation (2 changes)

    • #559 Deprecated functions and classes are now correctly visualized in the documentation.
    • #560 Updated development roadmap accordingly to 0.10, 0.11 and 0.12 releases.

    Deprecations (3 changes)

    • #532 Method CPreProcess.revert() is deprecated. Use .inverse_transform() instead.
    • #552 CClassifierKDE is now deprecated. Use CClassifierSkLearn with sklearn.neighbors.KernelDensity instead.
    • #553 CClassifierMCSLinear is now deprecated. Use CClassifierSkLearn with sklearn.ensemble.BaggingClassifier instead.
  • v0.9-rc2 protected
    0b88c4bc · Release v0.9-rc2 ·
  • v0.8.1 protected
    4d75ae4a · Release v0.8.1 ·
    Release v0.8.1


    This version does not contain any significant change.

    Documentation (2 changes)

    • #523 Fixed documentation not compiling under Sphinx v2.2.
    • #529 Updated roadmap accordingly for v0.9 release.
  • v0.8 protected
    22ca1462 · Release v0.8 ·
    Release v0.8


    • First public release!