ErfAct and PSerf: Non-monotonic smooth trainable Activation Functions.
Koushik BiswasSandeep KumarShilpak BanerjeeAshish Kumar PandeyPublished in: CoRR (2021)
Keyphrases
- activation function
- neural network
- feed forward
- artificial neural networks
- neural architecture
- hidden layer
- feed forward neural networks
- neural nets
- back propagation
- complex valued
- learning rate
- basis functions
- multilayer perceptron
- hidden nodes
- radial basis function
- network architecture
- fuzzy neural network
- training phase
- high dimensional