Two-argument activation functions learn soft XOR operations like cortical neurons.
KiJung YoonA. Emin OrhanJuhyun KimXaq PitkowPublished in: CoRR (2021)
Keyphrases
- activation function
- neural network
- hidden layer
- feed forward
- artificial neural networks
- receptive fields
- back propagation
- feed forward neural networks
- connection weights
- learning rate
- neural nets
- hidden neurons
- multilayer perceptron
- visual cortex
- network architecture
- neural architecture
- radial basis function
- basis functions
- recurrent neural networks
- biologically plausible
- hidden nodes
- spiking neurons
- fuzzy neural network
- training phase
- expert systems
- bayesian networks