Alpha Dropout is a type of dropout layer that maintains the mean and variance of the original inputs in order to ensure the self-normalizing property of SELU networks with dropout. Alpha Dropout fits with SELU networks by randomly setting activations to the negative saturation value of the activation function at a given ratio each pass.
Note: Alpha Dropout is generally only used in the context of SELU networks. Use regular Dropout for other types of neural nets.
|1||ratio||0.1||float||The ratio of neurons that are dropped during each training pass.|
use Rubix\ML\NeuralNet\Layers\AlphaDropout; $layer = new AlphaDropout(0.1);
- G. Klambauer et al. (2017). Self-Normalizing Neural Networks.