Sampling-based Gradient Regularization for Capturing Long-Term Dependencies in Recurrent Neural Networks.
Artem N. ChernodubDimitri NowickiPublished in: CoRR (2016)
Keyphrases
- recurrent neural networks
- long term
- short term
- neural network
- feed forward
- recurrent networks
- complex valued
- feedforward neural networks
- long short term memory
- reservoir computing
- cascade correlation
- echo state networks
- neural model
- monte carlo
- image gradient
- regularization parameter
- artificial neural networks
- edge detection
- real time
- chaotic time series
- reinforcement learning
- data mining
- em algorithm
- nonlinear dynamic systems
- genetic algorithm