KALE: When Energy-Based Learning Meets Adversarial Training.
Michael ArbelLiang ZhouArthur GrettonPublished in: CoRR (2020)
Keyphrases
- supervised learning
- learning algorithm
- reinforcement learning
- learning scheme
- feedforward neural networks
- learning process
- prior knowledge
- incremental learning
- learning problems
- online learning
- deep architectures
- multi agent
- e learning
- data sets
- online training
- error back propagation
- training phase
- higher education
- test set
- training samples
- active learning
- training set
- neural network