secure_learning.models.common_gradient_forms

Provides classes for computing the gradient of objective functions

Classes

GradientFunction	Class for objective function.
WeightedDifferencesGradient	Class for computing the gradient of objective functions. The gradient

Module Contents

class secure_learning.models.common_gradient_forms.GradientFunction

Class for objective function.

class secure_learning.models.common_gradient_forms.WeightedDifferencesGradient(predictive_func: Callable[[tno.mpc.mpyc.secure_learning.utils.Matrix[mpyc.sectypes.SecureFixedPoint], tno.mpc.mpyc.secure_learning.utils.Vector[mpyc.sectypes.SecureFixedPoint]], tno.mpc.mpyc.secure_learning.utils.Vector[mpyc.sectypes.SecureFixedPoint]])

Class for computing the gradient of objective functions. The gradient is assumed to have the following form: $$g(X, y, w) = X^T (f(X, w) - y)$$

We refer to $f$ as the predictive function.

predictive_func