Unreasonable effectiveness of learning neural networks: From accessible states and robust ensembles to basic algorithmic schemes.
Carlo BaldassiChristian BorgsJennifer T. ChayesAlessandro IngrossoCarlo LucibelloLuca SagliettiRiccardo ZecchinaPublished in: Proc. Natl. Acad. Sci. USA (2016)
Keyphrases
- neural network
- learning algorithm
- learning tasks
- learning process
- learning systems
- unsupervised learning
- connectionist systems
- feedforward neural networks
- neural nets
- prior knowledge
- artificial neural networks
- expert systems
- supervised learning
- knowledge acquisition
- fuzzy logic
- inductive inference
- learning rules
- reinforcement learning
- adaptive resonance theory
- neural network ensemble
- error back propagation
- machine learning