Time Delay Learning by Gradient Descent in Recurrent Neural Networks

Recurrent Neural Networks (RNNs) possess an implicit internal memory and are well adapted for time series forecasting. Unfortunately, the gradient descent algorithms which are commonly used for their training have two main weaknesses: the slowness and the difficulty of dealing with long-term dependencies in time series. Adding well chosen connections with time delays to the RNNs often reduces learning times and allows gradient descent algorithms to find better solutions. In this article, we demonstrate that the principle of time delay learning by gradient descent, although efficient for feed-forward neural networks and theoretically adaptable to RNNs, shown itself to be difficult to use in this latter case.

[1]  Jürgen Schmidhuber,et al.  Applying LSTM to Time Series Predictable through Time-Window Approaches , 2000, ICANN.

[2]  Yoshua Bengio,et al.  Learning long-term dependencies with gradient descent is difficult , 1994, IEEE Trans. Neural Networks.

[3]  Geoffrey E. Hinton,et al.  Learning internal representations by error propagation , 1986 .

[4]  Ah Chung Tsoi,et al.  Locally recurrent globally feedforward networks: a critical review of architectures , 1994, IEEE Trans. Neural Networks.

[5]  P. Gallinari,et al.  Recurrent Neural Networks with Delays , 1994 .

[6]  David E. Rumelhart,et al.  Predicting the Future: a Connectionist Approach , 1990, Int. J. Neural Syst..

[7]  Jean Pierre Asselin de Beauville,et al.  Learning long-term dependencies by the selective addition of time-delayed connections to recurrent neural networks , 2002, Neurocomputing.

[8]  Peter Tiño,et al.  Learning long-term dependencies in NARX recurrent neural networks , 1996, IEEE Trans. Neural Networks.

[9]  James L. McClelland,et al.  Parallel distributed processing: explorations in the microstructure of cognition, vol. 1: foundations , 1986 .

[10]  Romuald Boné,et al.  A bounded exploration approach to constructive algorithms for recurrent neural networks , 2000, Proceedings of the IEEE-INNS-ENNS International Joint Conference on Neural Networks. IJCNN 2000. Neural Computing: New Challenges and Perspectives for the New Millennium.

[11]  Richard J. Duro,et al.  Discrete-time backpropagation for training synaptic delay-based artificial neural networks , 1999, IEEE Trans. Neural Networks.

[12]  Barak A. Pearlmutter Dynamic recurrent neural networks , 1990 .

[13]  Ah Chung Tsoi,et al.  A unifying view of some training algorithms for multilayer perceptrons with FIR filter synapses , 1994, Proceedings of IEEE Workshop on Neural Networks for Signal Processing.

[14]  Bhaskar D. Rao,et al.  On-line learning algorithms for locally recurrent neural networks , 1999, IEEE Trans. Neural Networks.

[15]  L. Glass,et al.  Oscillation and chaos in physiological control systems. , 1977, Science.