Abstract
Generally, neural networks can be divided into two large classes. One class contains feedforward neural networks (FNNs), and the other contains recurrent neural networks (RNNs). This book focused on RNNs only. The essential difference between FNNs and RNNs is the presence of a feedback mechanism among the neurons in the latter. A FNN is a network without any feedback connections among its neurons, while a RNN has at least one feedback connection. Since RNNs allow feedback connections in neurons, the network topology can be very general: any neuron can be connected to any other, even to itself. Allowing the presence of feedback connections among neurons has an advantage, it leads naturally to an analysis of the networks as dynamic systems, in which the state of a network, at one moment in time, depends on the state at a previous moment in time. The topology of RNNs is shown in Figure 1.1.
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.
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2004 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Yi, Z., Tan, K.K. (2004). Introduction. In: Convergence Analysis of Recurrent Neural Networks. Network Theory and Applications, vol 13. Springer, Boston, MA. https://doi.org/10.1007/978-1-4757-3819-3_1
Download citation
DOI: https://doi.org/10.1007/978-1-4757-3819-3_1
Publisher Name: Springer, Boston, MA
Print ISBN: 978-1-4757-3821-6
Online ISBN: 978-1-4757-3819-3
eBook Packages: Springer Book Archive