Recurrent Neural Networks Training Using Derivative Free Nonlinear Bayesian Filters.
Branimir TodorovicMiomir S. StankovicClaudio MoragaPublished in: IJCCI (Selected Papers) (2014)
Keyphrases
- recurrent neural networks
- echo state networks
- derivative free
- recurrent networks
- nonlinear dynamic systems
- feedforward neural networks
- neural learning
- neural network
- gradient method
- unconstrained optimization
- feed forward
- optimization methods
- hidden layer
- neural model
- fan beam
- constrained optimization
- multi layered
- supervised learning
- objective function
- biologically inspired
- image reconstruction