Dynamic neural networks: an overview

Over the last decade several advances have been made in the paradigm of artificial neural networks with specific emphasis on architectures and learning algorithms. However, most of the work is focused on static (feedforward) neural networks. These neural networks respond instantaneously to the inputs, for they do not possess any time delay units. The use of time delays in neural networks is neurobiologically motivated, since it is well known that signal delays are omnipresent in the brain and play an important role in neurobiological information processing. This concept has led to the development of dynamic neural networks. It is envisaged that dynamic neural networks, in addition to better representation of biological neural systems, offer better computational capabilities compared to their static counterparts. The objective of this paper is to give an overview of dynamic neural structures.