On the Impact of the Activation Function on Deep Neural Networks Training.
Soufiane HayouArnaud DoucetJudith RousseauPublished in: CoRR (2019)
Keyphrases
- activation function
- neural network
- feed forward neural networks
- backpropagation algorithm
- single hidden layer
- feedforward neural networks
- back propagation
- hidden layer
- training phase
- artificial neural networks
- multi layer perceptron
- training algorithm
- feed forward
- multilayer perceptron
- neural nets
- learning rate
- hidden neurons
- extreme learning machine
- training process
- pattern recognition
- basis functions
- hidden nodes
- multi layer
- radial basis function
- output layer
- single layer
- network architecture
- machine learning
- hidden units
- recurrent neural networks
- linear combination
- fuzzy logic
- fuzzy neural network
- rbf neural network
- active learning
- training set
- learning algorithm
- sigmoid function