2002 WSEAS Int. Conf. on Signal Processing, Cádiz (España). 12-16 Junio 2002
Resumen:
This paper describes a new neural network able to adapt itself, both its parameters and its structure, to a data set in real-time conditions. The adaptation is based on a non-supervised learning procedure. The new neural network can automatically create interconnections between neurons using a Gaussian activation function. Still another important feature of this new neural network is the use of few neurons to make a good prediction using a reduced number of examples. This is relevant in order to make fast calculations using few resources in real-time applications. Some examples focusing on mobile robotics applications are included in order to demonstrate its good performance.
Palabras clave: one-pass learning, neural networks, environment modeling, real-time navigation, autonomous, mobile robot.
Fecha de publicación: junio 2002.
Cita:
Sánchez, A., Sanz-Bobi, M.A., Real Time Dynamic Ellipsoidal Neural Network (RTDENN), 2002 WSEAS Int. Conf. on Signal Processing, Cádiz (España). 12-16 Junio 2002.
IIT-02-036A