NeuralNet v1.0.0 NeuralNet.ActivationFunctions

A few activation functions and their derivatives.

Summary

This sigmoid has a range of 0 to 1. It uses 1 / (1 + e^-x)

Returns the derivative of this sigmoid function at x

tanh(x)

Tanh has a range of -1 to 1

Returns the derivative of tanh at x

sigmoid(x)

This sigmoid has a range of 0 to 1. It uses 1 / (1 + e^-x)

sigmoid_prime(x)

Returns the derivative of this sigmoid function at x.

tanh(x)

Tanh has a range of -1 to 1.

tanh_prime(x)

Returns the derivative of tanh at x.