Abstract
In this paper a critical review of gradient-based training methods for recurrent neural networks is presented including Back Propagation Through Time (BPTT), Real Time Recurrent Learning (RTRL) and several specific learning algorithms for different locally recurrent architectures. From this survey it comes out the need for a unifying view of all the specific procedures proposed for networks with local feedbacks, that keeps into account the general framework of recurrent networks learning: BPTT and RTRL. Therefore a learning method for local feedback network is proposed which combines together the best feature of BPTT, i.e. the lowest complexity, and of RTRL, i.e. the on-line operation, and includes as special case several specific algorithms already proposed, such as Temporal Back Propagation, Back Propagation for Sequences, Back-Tsoi algorithm and some others. In the general version, this new training method allows on-line efficient and accurate gradient calculation. It compares favourably with the previous algorithms in stability, speed/complexity trade off, accuracy.
Keywords
- Back Propagation
- Finite Impulse Response
- Recurrent Neural Network
- Multi Layer Perceptron
- Neural Computation
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
This research was supported by the Italian MURST.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
R.J. Williams, J. Peng. An efficient gradient-based algorithm for on line training of recurrent network trajectories. Neural Computation 2: 490–501,1990.
R.J. Williams, D. Zipser. A Learning Algorithm for Continually Running Fully Recurrent Neural Networks. Neural Computation 1: 270–280, 1989.
A.D. Back, A.C. Tsoi. FIR and IIR synapses, a new neural network architecture for time series modelling. Neural Computation 3: 375–385, 1991.
A.C. Tsoi, A.D. Back. Locally recurrent globally feedforward networks: a critical review of architectures. IEEE Transactions on Neural Networks, vol. 5, no. 2, 229–239, March 1994.
P.J. Werbos. Beyond regression: New tools for prediction and analysis in the behavioural sciences. Ph.D. dissertation, Committee on Appl. Math., Harvard Univ., Cambridge, MA, Nov. 1974.
P.J. Werbos. Backpropagation through time: what it does and how to do it. Proceedings of IEEEa Special issue on neural networks, vol. 78, No. 10, pp. 1550–1560, October 1990.
E.A. Wan. Temporal backpropagation for FIR neural networks. Proceedings of the International Joint Conference on Neural Networks, 1: 575–580, 1990.
A. Waibel, T. Hanazawa, G. Hinton, K. Shikano, K.J. Lang. Phoneme recognition using time-delay neural networks. IEEE Trans, on Acoustic, Speech, and Signal Processing, Vol. 37, No. 3, March 1989.
N. Benvenuto, F. Piazza, A. Uncini. Comparison of four learning algorithms for multilayer perceptron with FIR synapses. Proceeding of the IEEE International Conference of Neural Networks, 1994.
J.J. Shynk. Adaptive IIR filtering. IEEE ASSP Magazine, April 1989.
P. Campolucci, F. Piazza, A. Uncini. On-line learning algorithms for neural networks with IIR synapses. Proc. of the IEEE International Conference of Neural Networks, Perth, Nov. 1995.
P. Campolucci, A. Uncini, F. Piazza. Causal Back Propagation Through Time for Locally Recurrent Neural Networks. Proc. of the IEEE International Symposium on Circuits and Systems, Atlanta, May 1996.
Y.Bengio, R. De Mori, M.Gori. Learning the dynamic of speech with back-propagation for sequences. Pattern Recognition Letters, 13: 375–385, 1992.
P.Frasconi, M.Gori, G.Soda. Local Feedback Multilayered Networks. Neural Computation 4: 120–130, 1992.
B.A.Pearlmutter. Gradient Calculations for Dynamic Recurrent Neural Networks: A Survey. IEEE Trans, on Neural Networks vol. 6, no. 5, September 1995.
B. Srinivasan, U.R. Prasad and N.J. Rao. Backpropagation through Adjoints for the identification of Non linear Dynamic Systems using Recurrent Neural Models. IEEE Trans, on Neural Networks pp. 213–228, March 1994.
E.A.Wan, F. Beaufays. Diagrammatic Derivation of Gradient Algorithms for Neural Networks. Neural Computation 8: 182–201, 1996.
K.S. Narendra, K. Parthasarathy. Identification and control of dynamical systems using neural networks. IEEE Trans, on Neural Networks, vol. 1, pp. 4–27, March 1990.
C.-C. Ku, K.Y.Lee. Diagonal Recurrent Neural Networks for Dynamic Systems Control. IEEE Trans, on Neural Networks vol. 6, no. 1, January 1995.
F.Beaufays, E.Wan. Relating Real-Time Backpropagation and Backpropagation-Through-Time: An Application of Flow Graph Interreciprocity. Neural Computation 6: 296–306, 1994.
A.D.Back, E.Wan, S.Lawrence, A.C.Tsoi. A unifying view of some training algorithms for multilayer perceptrons with FIR filter synapses. Proc. EEEE workshop on Neural Networks for Signal Processing, pp. 146–154, 1994.
B.A. Pearlmutter. Two New Learning Procedures for Recurrent Networks. Neural Networks Rev. vol. 3, no. 3, pp. 99–101, 1990.
J.L. Elman. Finding Structure in Time. Cognitive Science 14: 179–211, 1990.
M.C. Mozer. A Focused Back-propagation Algorithm for Temporal Pattern Recognition. Tech Rep. CRG-TR-88-3, University of Toronto, 1988 and Complex Systems 3: 349–381, 1989.
R.R. Leighton and B.C. Conrath. The Autoregressive Backpropagation Algorithm. Proc. International Joint Conference on Neural Networks, pp. 369–377, 1991.
L.B. Almeida. A learning rule for asynchronous perceptrons with feedback in combinatorial environment. Proc. International Conference on Neural Networks, vol. 2, pp. 609–618, 1987.
R.J. Williams, D. Zipser. Gradient-based learning algorithms for recurrent networks and their computational complexity. In Backpropagation: Theory, Architectures and Applications Y. Chauvin and D.E. Rumelhart, Eds. Hillsdale, NJ: Lawrence Erlbaum Associates, 1994.
T. Uchiyama, K. Shimohara, Y. Tokunaga. A modified leaky integrator network for temporal pattern recognition. Proc. of the International Joint Conference on Neural Networks, vol. 1, pp. 469–475, 1989.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1998 Springer-Verlag London Limited
About this paper
Cite this paper
Campolucci, P., Uncini, A., Piazza, F. (1998). A Unifying View of Gradient Calculations and Learning for Locally Recurrent Neural Networks. In: Marinaro, M., Tagliaferri, R. (eds) Neural Nets WIRN VIETRI-97. Perspectives in Neural Computing. Springer, London. https://doi.org/10.1007/978-1-4471-1520-5_3
Download citation
DOI: https://doi.org/10.1007/978-1-4471-1520-5_3
Publisher Name: Springer, London
Print ISBN: 978-1-4471-1522-9
Online ISBN: 978-1-4471-1520-5
eBook Packages: Springer Book Archive