Title :
Progressive learning algorithm for GSN feedforward neural architectures
Author :
de Carvalho, A. ; Fairhurst, Michael C.
Author_Institution :
Electron. Eng. Labs., Kent Univ., Canterbury
fDate :
3/17/1994 12:00:00 AM
Abstract :
A new learning algorithm, the progressive learning algorithm, is proposed for use in a goal-seeking feedforward architecture, GSNf . Experiments show that both improved recognition performance and significantly faster training times can be achieved with this learning algorithm
Keywords :
feedforward neural nets; learning (artificial intelligence); pattern recognition; feedforward neural architectures; goal-seeking neuron; progressive learning algorithm; recognition performance; training times;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:19940326
