Use relu function for lstmlayer

I would like to change the StateActivationFuction of lstmLayer to Relu fuction, but only 'tanh' and 'softsign' are supported in the deep learning tool box.

Is there any solutions for changing the activation function ,or the way to make customed lstmLayer with Relu as the StateActivation?

Answer 1

[slevin Lee]

GateActivationFunction — Activation function to apply to the gates

• 'sigmoid' – Use the sigmoid function σ(x)=(1+e−x)−1.
• 'hard-sigmoid' – Use the hard sigmoid function

no Relu fuction

╮(╯▽╰)╭