In this paper we present a two layer neural network architecture based on the Adaptive Resonance Theory (ART) paradigm. It merges the unsupervised training capability of an ART-like structure with the supervised MLP algorithm by adapting the selective attention mechanisms to the pattern-class characteristics. It has been successfully applied to pattern recognition tasks such as isolate word recognition.
SARLA: a supervised adaptive resonance learning algorithm: principles and applications
V Rampa
1993
Abstract
In this paper we present a two layer neural network architecture based on the Adaptive Resonance Theory (ART) paradigm. It merges the unsupervised training capability of an ART-like structure with the supervised MLP algorithm by adapting the selective attention mechanisms to the pattern-class characteristics. It has been successfully applied to pattern recognition tasks such as isolate word recognition.File in questo prodotto:
Non ci sono file associati a questo prodotto.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
