Abstract. An architecture for achieving word recognition and incremental learning of new words in a language processing system is presented. The architecture is based on neural ass...
Insufficient training data is one of the major problems in neural network learning, because it leads to poor learning performance. In order to enhance an intelligent learning proc...
The self-organizing map SOM is widely used as a data visualization method in various engineering applications. It performs a non-linear mapping from a high-dimensional data spac...
The well-known backpropagation (BP) derivative computation process for multilayer perceptrons (MLP) learning can be viewed as a simplified version of the Kelley-Bryson gradient f...
Multicast conferencing is a rapidly-growing area of Internet use. Audio, video and other media such as shared whiteboard data can be distributed efficiently between groups of confe...
Mourad Amad, Zahir Haddad, Lachemi Khenous, Kamal ...