Variational Autoencoders

DeepLearningMachineLearningStatistics A Variational autoencoder (VAE) is a Latent Variable Model that assumes the data is generated from a normal distribution whose parameters are learned by neural networks. The Bayesian Network, has joint factorization where