Stabilizing Training of Generative Adversarial Networks through Regularization.
Kevin RothAurélien LucchiSebastian NowozinThomas HofmannPublished in: CoRR (2017)
Keyphrases
- echo state networks
- training process
- early stopping
- training set
- data driven
- generative model
- discriminative training
- recurrent networks
- training phase
- training algorithm
- training samples
- training examples
- stochastic gradient descent
- regularization parameter
- complex networks
- online learning
- social networks
- test set
- neural network
- prior information
- maximum likelihood
- supervised learning
- probabilistic model
- hidden markov models
- bayesian networks
- training data
- labelled data
- e learning