Why Do Deep Neural Networks with Skip Connections and Concatenated Hidden Representations Work?
Oyebade K. OyedotunDjamila AouadaPublished in: ICONIP (3) (2020)
Keyphrases
- neural network
- distributed representations
- pattern recognition
- neural nets
- multilayer perceptron
- higher level
- feature vectors
- genetic algorithm
- deep belief networks
- fuzzy logic
- back propagation
- network architecture
- deep learning
- competitive learning
- multi layer
- symbolic knowledge
- neural network model
- recurrent neural networks
- feed forward
- self organizing maps
- machine learning
- real time
- symbolic representation
- website
- learning algorithm
- data mining
- unsupervised feature learning