An Efficient Gradient-Based Algorithm for On-Line Training of Recurrent Network Trajectories
Ronald J. Williams, Jing Peng · Neural Computation · 1990
A novel variant of the familiar backpropagation-through-time approach to training recurrent networks is described. This algorithm is intended to be used on arbitrary recurrent networks that run continually without ever being reset to an initial state, and it is specifically designed for computationally efficient computer implementation. This algorithm can be viewed as a cross between epochwise backpropagation through time, which is not appropriate for continually running networks, and the widely used on-line gradient approximation technique of truncated backpropagation through time.
Cited by 577 works. doi:10.1162/neco.1990.2.4.490
Authors
References
- Learning Internal Representations by Error Propagation
- A Learning Algorithm for Continually Running Fully Recurrent Neural Networks
- Generalization of back-propagation to recurrent neural networks
- Learning State Space Trajectories in Recurrent Neural Networks
- A learning rule for asynchronous perceptrons with feedback in a combinatorial environment
- Finite State Automata and Simple Recurrent Networks
- Experimental Analysis of the Real-time Recurrent Learning Algorithm
- A learning algorithm for analog, fully recurrent neural networks
- A Subgrouping Strategy that Reduces Complexity and Speeds Up Learning in Recurrent Networks
- Comparison of four neural net learning methods for dynamic system identification