Training Very Deep Networks via Residual Learning with Stochastic Input Shortcut Connections.
Oyebade K. OyedotunAbd El Rahman ShabayekDjamila AouadaBjörn E. OtterstenPublished in: ICONIP (2) (2017)
Keyphrases
- learning process
- learning systems
- recurrent networks
- learning algorithm
- online learning
- prior knowledge
- reinforcement learning
- learning tasks
- learning problems
- active learning
- training process
- connectionist networks
- deep architectures
- supervised learning
- knowledge acquisition
- unsupervised learning
- feedforward neural networks
- learning speed
- machine learning
- online training
- computer based training