Using Self-Supervised Learning Can Improve Model Robustness and Uncertainty.
Dan HendrycksMantas MazeikaSaurav KadavathDawn SongPublished in: NeurIPS (2019)
Keyphrases
- learning scheme
- learning models
- statistical model
- computational model
- parameter estimation
- formal model
- prior knowledge
- probabilistic model
- automatically learned
- knowledge acquisition
- probability distribution
- topic models
- active learning
- mathematical model
- artificial neural networks
- experimental data
- network structure
- conceptual model
- learning scenarios
- reinforcement learning
- decision trees
- neural network
- inductive inference
- learning phase
- data sets