Gaussian Process Neurons Learn Stochastic Activation Functions.
Sebastian UrbanMarcus BasallaPatrick van der SmagtPublished in: CoRR (2017)
Keyphrases
- activation function
- gaussian process
- hidden layer
- neural network
- latent variables
- gaussian processes
- artificial neural networks
- feed forward
- hyperparameters
- neural nets
- bayesian framework
- learning rate
- multilayer perceptron
- model selection
- regression model
- back propagation
- semi supervised
- basis functions
- radial basis function
- fuzzy neural network
- training phase
- recurrent neural networks
- data sets
- network architecture
- multi layer perceptron
- prior knowledge
- noise level
- cross validation
- pairwise
- rbf neural network
- em algorithm
- probabilistic model