Application-Driven Neural Networks

Appliaction-driven neural networks are loosely tied to the biological realities. As the knowledge on nervous systems is not advanced, we have to define different functionalities and connection structures apart from a biological perspective. The strength of application-driven neural networks hinges upon three main characteristics:
  1. Adaptiveness and self-organization: it offers robust and adaptive processing capabilities by adopting adaptive learning and self-organization rules.
  2. Nonlinear network processing: it enhances the network's aproximation, classification and noise-inmunity capabilities.
  3. Parallel processing: it usually employs a large number of processing cells enhanced by extensive interconnectivity.
These characteristics have played an important role in ANNs, applicabilities to signal and image processing and analysis. An application driven neural model can be very precisely defined. A NN architecture comprises massively parallel adaptive processing elements with hierarchically structured interconnection networks.

Taxonomy of Neural Networks

Contents

Artificial Neural Networks
About this Tutorial