alibi.explainers.cem

Constants

DEFAULT_DATA_CEM

DEFAULT_DATA_CEM: dict = {'PN': None, 'PP': None, 'PN_pred': None, 'PP_pred': None, 'grads_graph': Non...

DEFAULT_META_CEM

DEFAULT_META_CEM: dict = {'name': None, 'type': ['blackbox', 'tensorflow', 'keras'], 'explanations': [...

logger

logger: logging.Logger = <Logger alibi.explainers.cem (WARNING)>

Instances of the Logger class represent a single logging channel. A "logging channel" indicates an area of an application. Exactly how an "area" is defined is up to the application developer. Since an application can have any number of areas, logging channels are identified by a unique string. Application areas can be nested (e.g. an area of "input processing" might include sub-areas "read CSV files", "read XLS files" and "read Gnumeric files"). To cater for this natural nesting, channel names are organized into a namespace hierarchy where levels are separated by periods, much like the Java or Python package namespace. So in the instance given above, channel names might be "input" for the upper level, and "input.csv", "input.xls" and "input.gnu" for the sub-levels. There is no arbitrary limit to the depth of nesting.

CEM

Inherits from: Explainer, FitMixin, ABC, Base

Constructor

CEM(self, predict: Union[Callable[[numpy.ndarray], numpy.ndarray], keras.src.models.model.Model], mode: str, shape: tuple, kappa: float = 0.0, beta: float = 0.1, feature_range: tuple = (-10000000000.0, 10000000000.0), gamma: float = 0.0, ae_model: Optional[keras.src.models.model.Model] = None, learning_rate_init: float = 0.01, max_iterations: int = 1000, c_init: float = 10.0, c_steps: int = 10, eps: tuple = (0.001, 0.001), clip: tuple = (-100.0, 100.0), update_num_grad: int = 1, no_info_val: Union[float, numpy.ndarray, NoneType] = None, write_dir: Optional[str] = None, sess: Optional[tensorflow.python.client.session.Session] = None) -> None

Name	Type	Default	Description
predict	Union[Callable[[.[<class 'numpy.ndarray'>]], numpy.ndarray], keras.src.models.model.Model]		tensorflow model or any other model's prediction function returning class probabilities.
mode	str		Find pertinent negatives (PN) or pertinent positives (PP).
shape	tuple		Shape of input data starting with batch size.
kappa	float	0.0	Confidence parameter for the attack loss term.
beta	float	0.1	Regularization constant for L1 loss term.
feature_range	tuple	(-10000000000.0, 10000000000.0)	Tuple with min and max ranges to allow for perturbed instances. Min and max ranges can be float or numpy arrays with dimension (1x nb of features) for feature-wise ranges.
gamma	float	0.0	Regularization constant for optional auto-encoder loss term.
ae_model	Optional[keras.src.models.model.Model]	None	Optional auto-encoder model used for loss regularization.
learning_rate_init	float	0.01	Initial learning rate of optimizer.
max_iterations	int	1000	Maximum number of iterations for finding a PN or PP.
c_init	float	10.0	Initial value to scale the attack loss term.
c_steps	int	10	Number of iterations to adjust the constant scaling the attack loss term.
eps	tuple	(0.001, 0.001)	If numerical gradients are used to compute dL/dx = (dL/dp) * (dp/dx), then eps[0] is used to calculate dL/dp and eps[1] is used for dp/dx. eps[0] and eps[1] can be a combination of float values and numpy arrays. For eps[0], the array dimension should be (1x nb of prediction categories) and for eps[1] it should be (1x nb of features).
clip	tuple	(-100.0, 100.0)	Tuple with min and max clip ranges for both the numerical gradients and the gradients obtained from the tensorflow graph.
update_num_grad	int	1	If numerical gradients are used, they will be updated every update_num_grad iterations.
no_info_val	Union[float, numpy.ndarray, None]	None	Global or feature-wise value considered as containing no information.
write_dir	Optional[str]	None	Directory to write tensorboard files to.
sess	Optional[tensorflow.python.client.session.Session]	None	Optional tensorflow session that will be used if passed instead of creating or inferring one internally.

Methods

attack

attack(X: numpy.ndarray, Y: numpy.ndarray, verbose: bool = False) -> Tuple[numpy.ndarray, Tuple[numpy.ndarray, numpy.ndarray]]

Name	Type	Default	Description
X	numpy.ndarray		Instance to attack.
Y	numpy.ndarray		Labels for X.
verbose	bool	False	Print intermediate results of optimization if True.

Returns

• Type: Tuple[numpy.ndarray, Tuple[numpy.ndarray, numpy.ndarray]]

explain

explain(X: numpy.ndarray, Y: Optional[numpy.ndarray] = None, verbose: bool = False) -> alibi.api.interfaces.Explanation

Name	Type	Default	Description
X	numpy.ndarray		Instances to attack.
Y	Optional[numpy.ndarray]	None	Labels for X.
verbose	bool	False	Print intermediate results of optimization if True.

Returns

• Type: alibi.api.interfaces.Explanation

fit

fit(train_data: numpy.ndarray, no_info_type: str = 'median') -> alibi.explainers.cem.CEM

Name	Type	Default	Description
train_data	numpy.ndarray		Representative sample from the training data.
no_info_type	str	'median'	Median or mean value by feature supported.

Returns

• Type: alibi.explainers.cem.CEM

get_gradients

get_gradients(X: numpy.ndarray, Y: numpy.ndarray) -> numpy.ndarray

Name	Type	Default	Description
X	numpy.ndarray		Instance around which gradient is evaluated.
Y	numpy.ndarray		One-hot representation of instance labels.

Returns

• Type: numpy.ndarray

loss_fn

loss_fn(pred_proba: numpy.ndarray, Y: numpy.ndarray) -> numpy.ndarray

Name	Type	Default	Description
pred_proba	numpy.ndarray		Prediction probabilities of an instance.
Y	numpy.ndarray		One-hot representation of instance labels.

Returns

• Type: numpy.ndarray

perturb

perturb(X: numpy.ndarray, eps: Union[float, numpy.ndarray], proba: bool = False) -> Tuple[numpy.ndarray, numpy.ndarray]

Name	Type	Default	Description
X	numpy.ndarray		Array to be perturbed.
eps	Union[float, numpy.ndarray]		Size of perturbation.
proba	bool	False	If True, the net effect of the perturbation needs to be 0 to keep the sum of the probabilities equal to 1.

Returns

• Type: Tuple[numpy.ndarray, numpy.ndarray]

reset_predictor

reset_predictor(predictor: Union[Callable, keras.src.models.model.Model]) -> None

Name	Type	Default	Description
predictor	Union[Callable, keras.src.models.model.Model]		New predictor function/model.

Returns

• Type: None