Training Recurrent Neural Networks as a Constraint Satisfaction Problem.
Hamid KhodabandehlouMohammad Sami FadaliPublished in: CoRR (2018)
Keyphrases
- recurrent neural networks
- constraint satisfaction problems
- recurrent networks
- echo state networks
- constraint satisfaction
- feedforward neural networks
- constraint propagation
- feed forward
- neural network
- constraint programming
- artificial neural networks
- non binary
- np complete
- constraint solving
- np hard
- constraint networks
- search space
- partial constraint satisfaction
- reservoir computing
- cascade correlation
- finite domain
- soft constraints
- nonlinear dynamic systems
- product configuration
- constraint solver
- arc consistency
- reasoning problems
- disjunctive constraints
- multi layer perceptron
- disjunctive temporal
- temporal constraints
- forward checking
- hard constraints
- distributed constraint satisfaction
- decomposition methods
- fuzzy logic
- pseudo boolean optimization