Widely Linear Kernels for Complex-Valued Kernel Activation Functions.
Simone ScardapaneSteven Van VaerenberghDanilo ComminielloAurelio UnciniPublished in: CoRR (2019)
Keyphrases
- complex valued
- activation function
- kernel function
- kernel methods
- neural network
- real valued
- feed forward
- artificial neural networks
- back propagation
- feature space
- hidden layer
- neural nets
- input space
- support vector
- recurrent neural networks
- learning rate
- multilayer perceptron
- linear equations
- radial basis function
- network architecture
- basis functions
- training phase
- learning tasks
- support vector machine svm
- linear combination
- signal processing
- fuzzy logic
- fuzzy neural network
- support vector machine
- linear systems
- active learning
- machine learning