An Efficient Gradient-Based Algorithm for On-Line Training of Recurrent Network Trajectories.
Ronald J. WilliamsJing PengPublished in: Neural Comput. (1990)
Keyphrases
- times faster
- learning algorithm
- training algorithm
- computationally efficient
- probabilistic model
- high accuracy
- worst case
- similarity measure
- artificial neural networks
- optimal solution
- recurrent networks
- decision making
- matching algorithm
- optimization algorithm
- bayesian networks
- expectation maximization
- input data
- computational complexity
- neural network
- experimental evaluation
- cost function
- significant improvement
- k means
- linear programming
- supervised learning
- np hard
- recognition algorithm
- search space