rdkit.ML.Neural.ActFuncs module

Activation functions for neural network nodes

Activation functions should implement the following API:

• _Eval(x)_: returns the value of the function at a given point
• _Deriv(x)_: returns the derivative of the function at a given point

The current Backprop implementation also requires:

• _DerivFromVal(val)_: returns the derivative of the function when its

  value is val

In all cases _x_ is a float as is the value returned.

class rdkit.ML.Neural.ActFuncs.ActFunc

Bases: object

“virtual base class” for activation functions

class rdkit.ML.Neural.ActFuncs.Sigmoid(beta=1.0)

Bases: rdkit.ML.Neural.ActFuncs.ActFunc

the standard sigmoidal function

Deriv(x)

DerivFromVal(val)

Eval(x)

class rdkit.ML.Neural.ActFuncs.TanH(beta=1.0)

Bases: rdkit.ML.Neural.ActFuncs.ActFunc

the standard hyperbolic tangent function

Deriv(x)

DerivFromVal(val)

Eval(x)