Package org.ai_heuristic.functs.activate

Implementations of known activation functions that measure a firing limit threshold, but may work with other number types, through the org.ai_heuristic.eval.metric data structures.

See: Description

Class Summary

Class	Description
FunctionHardLimit	This function returns 1.0 of the value is larger than a threshold and 0.0 otherwise.
FunctionLinear	This function simply returns the first value in the dataset, as is, without any changes.
FunctionLinearPlus	This function returns a linear value calculated by y = (a * x) + b, where a and b are parameters of type double.
FunctionLinearRectified	This function simply returns the same first positive value in the dataset, as is, without any changes, or 0 if it is negative.
FunctionSigmoid	This function returns the sigmoid value as defined by (1 / (1 + e^-x)) where x is the input value and probably of type double.
FunctionSymmetricHardLimit	This function returns 1.0 of the value is larger than a threshold and -1.0 otherwise.
FunctionTanH	This function returns the tanh value as defined by ((e^2x - 1) / (e^2x + 1)) where x is the input value and probably of type double.

Package org.ai_heuristic.functs.activate Description

Implementations of known activation functions that measure a firing limit threshold, but may work with other number types, through the org.ai_heuristic.eval.metric data structures.