opendataval.model.GradientModel

class opendataval.model.GradientModel

Provides access to gradients of a Model

TODO Some data evaluators may benefit from higher-order gradients or hessians.

__init__()

Methods

__init__()
clone()	Clone Model object.
fit(x_train, y_train, *args[, sample_weights])	Fits the model on the training data.
grad(x_data, y_train, *args, **kwargs)	Given input data, iterates through the computed gradients of the model.
predict(x, *args, **kwargs)	Predict the label from the input covariates data.

Attributes

Models