Generalization in cascade-correlation networks

Author

Sjøgaard, Steen

Author_Institution

Dept. of Comput. Sci., Aarhus Univ., Denmark

fYear

1992

fDate

31 Aug-2 Sep 1992

Firstpage

59

Lastpage

68

Abstract

Two network construction algorithms are analyzed and compared theoretically as well as empirically. The first algorithm is the cascade correlation learning architecture proposed by S. E. Fahlman (1990), while the other algorithm is a small but striking modification of the former. Fahlman´s algorithm builds multilayer feedforward networks with as many layers as the number of added hidden units, while the other algorithm operates with just one layer of hidden units. This implies that their computational capabilities and the representation of the generalizations they deal with are quite diverse, and it is demonstrated how the generalization ability of the networks generated by Fahlman´s algorithm is outperformed by the networks built by the new algorithm

Keywords

generalisation (artificial intelligence); learning (artificial intelligence); neural nets; Fahlman´s algorithm; cascade correlation learning architecture; cascade-correlation networks; generalizations; network construction algorithms; neural nets; Algorithm design and analysis; Buildings; Computer architecture; Computer networks; Computer science; Computer vision; Education; Electronic mail; Intelligent networks; Neural networks;

fLanguage

English

Publisher

ieee

Conference_Titel

Neural Networks for Signal Processing [1992] II., Proceedings of the 1992 IEEE-SP Workshop

Conference_Location

Helsingoer

Print_ISBN

0-7803-0557-4

Type

conf

DOI

10.1109/NNSP.1992.253707

Filename

253707