The ReconstructionDistribution is used with variational autoencoders
Bernoulli reconstruction distribution for variational autoencoder.
Outputs are modelled by a Bernoulli distribution - i.e., the Bernoulli distribution should be used for binary data (all values 0 or 1); the VAE models the probability of the output being 0 or 1.
Consequently, the sigmoid activation function should be used to bound activations to the range of 0 to 1.
CompositeReconstructionDistribution is a reconstruction distribution built from multiple other ReconstructionDistribution instances.
The typical use is to combine for example continuous and binary data in the same model, or to combine different distributions for continuous variables.
Exponential reconstruction distribution.
Supports data in range [0,infinity)
Gaussian reconstruction distribution for variational autoencoder.
Outputs are modelled by a Gaussian distribution, with the mean and variances (diagonal covariance matrix) for each output determined by the network forward pass.
LossFunctionWrapper allows training of a VAE model with a standard (possibly deterministic) neural network loss function for the reconstruction, instead of using a
Variational Autoencoder layer