Abstract
Long Short-Term Memory (LSTM) Recurrent Neural networks (RNN) are improved and expanded architectures designed to overcome the shortcomings of training simple RNN (sRNN) simple RNN. They rely on three separate gating signals, each replicates a simple RNN, joined together by the ”memory-cell” structure. This architecture has been the effective and successful workhorse of RNN in numerous applications following its publication in 1997. The chapter introduces the standard architecture which embeds four replica of the simple RNN and thus has four times the parameters. Then, the chapter presents reduced-parameter in the gating signals by introducing five variants that maintain the gating structure while progressively reducing parameters. These five variants belong to the Slim LSTMs. The chapter ends with example case studies of the powerful standard LSTM RNN and the five slim LSTMs exhibiting their comparative performance.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Ahmad, M., & Salem, F. M. (1992). Dynamic learning using exponential energy functions. In Proceedings of the IEEE International Joint Conference on Neural Networks pp. (II–121–II–126).
Akandeh, A., & Salem, F. M. (2017a). Simplified long short-term memory recurrent neural networks: part I. arXiv:1707.04619.
Akandeh, A., & Salem, F. M. (2017b). Simplified long short-term memory recurrent neural networks: Part II. arXiv:1707.04623.
Akandeh, A., & Salem, F. M. (2017c). Simplified long short-term memory recurrent neural networks: Part III. arXiv:1707.04626.
Bengio, Y., Simard, P., & Frasconi, P. (1994). Learning long-term dependencies with gradient descent is difficult. IEEE Transactions on Neural Networks, 5(2), 157–166.
Chollet, F. Keras. https://keras.io
Chollet, F. Keras-codes2. https://github.com/keras-team/keras
Chollet, F. Keras-codes. https://github.com/keras-team/keras
Chung, J., Gulcehre, C., Cho, K., & Bengio, Y. (2014b). Empirical evaluation of gated recurrent neural networks on sequence modeling. arXiv preprint arXiv:1412.3555
Gers, F. A., Schraudolph, N. N., & Schmidhuber, J. (2002). Learning precise timing with LSTM recurrent networks. Journal of Machine Learning Research, 3, 115–143.
Graves, A. (2012). Supervised sequence labelling with recurrent neural networks. In Studies in Computational Intelligence. Springer.
Greff, K., Srivastava, R. K., Koutník, J., Steunebrink, B. R., & Schmidhuber, J. (2017). LSTM: A search space odyssey. IEEE Transactions on Neural Networks and Learning Systems, 28(10), 2222–2232.
Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural Computation, 9(8), 1735–1780.
Jianxin, C.-L., Zhang, G.-B., Zhou, Wu, J., & Zhou, Z.-H. (2016). Minimal gated unit for recurrent neural networks. https://arxiv.org/abs/1603.09420.
Johnson, M., Schuster, M., Le, Q. V., Krikun, M., Wu, Y., Chen, Z., Thorat, N., Viégas, F. B., Wattenberg, M., Corrado, G., M. Hughes, & Dean, J. (2016). Google’s multilingual neural machine translation system: Enabling zero-shot translation. http://arxiv.org/abs/1611.04558.
Le, Q. V., Jaitly, N., & Hinton, G. E. (2015). A simple way to initialize recurrent networks of rectified linear units. arXiv preprint arXiv:1504.00941
Lu, Y., & Salem, F. M. (2017). Simplified gating in long short-term memory (LSTM) recurrent neural networks. In 2017 IEEE 60th International Midwest Symposium on Circuits and Systems (MWSCAS) (p. 1601).
Salem, F. M. (2016a). A basic recurrent neural network model. arXiv preprint arXiv:1612.09022.
Salem, F. M. (2016b). Reduced parameterization in gated recurrent neural networks. Technical Report 11-2016, MSU.
Salem, F. M. (2018). Slim LSTMs. https://arxiv.org/abs/1812.11391
Zaremba, W. (2015). An empirical exploration of recurrent network architectures. In An empirical exploration of recurrent network architectures.
Zhou, G.-B., Wu, J., Zhang, C.-L., & Zhou, Z.-H. (2016). Minimal gated unit for recurrent neural networks. International Journal of Automation and Computing, 13(3), 226–234
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this chapter
Cite this chapter
Salem, F.M. (2022). Gated RNN: The Long Short-Term Memory (LSTM) RNN. In: Recurrent Neural Networks. Springer, Cham. https://doi.org/10.1007/978-3-030-89929-5_4
Download citation
DOI: https://doi.org/10.1007/978-3-030-89929-5_4
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-89928-8
Online ISBN: 978-3-030-89929-5
eBook Packages: EngineeringEngineering (R0)