About the Equivalence Between Complex-Valued and Real-Valued Fully Connected Neural Networks - Application to Polinsar Images.
Jose Agustin BarrachinaChenfang RenGilles VieillardChristèle MorisseauJean Philippe OvarlezPublished in: MLSP (2021)
Keyphrases
- real valued
- complex valued
- activation function
- neural network
- fully connected
- recurrent neural networks
- artificial neural networks
- feed forward
- latent variables
- ranking functions
- radial basis function
- similarity measure
- fault diagnosis
- multilayer perceptron
- segmentation method
- learning rate
- graphical models
- fuzzy logic