Cubature Kalman Filter Based Training of Hybrid Differential Equation Recurrent Neural Network Physiological Dynamic Models.
Ahmet DemirkayaTales ImbiribaKyle LockwoodSumientra RampersadElie AlhajjarGiovanna GuidoboniZachary DanzigerDeniz ErdogmusPublished in: CoRR (2021)
Keyphrases
- dynamic model
- recurrent neural networks
- differential equations
- recurrent networks
- feedforward neural networks
- echo state networks
- experimental data
- neural network structure
- dynamical systems
- neural network
- feed forward
- boundary value problem
- numerical methods
- complex valued
- initial conditions
- hidden layer
- numerical solution
- artificial neural networks
- reservoir computing
- continuous functions
- neural model
- decision making
- training process
- difference equations
- training phase
- ordinary differential equations
- multiple models
- control scheme
- partial differential equations
- back propagation
- higher order
- computer vision