NEURAL NETWORKS WITH QUADRATIC VC DIMENSION

Authors
KOIRAN P SONTAG ED
Citation
P. Koiran et Ed. Sontag, NEURAL NETWORKS WITH QUADRATIC VC DIMENSION, Journal of computer and system sciences, 54(1), 1997, pp. 190-198
Citations number
17
Categorie Soggetti
System Science","Computer Science Hardware & Architecture","Computer Science Theory & Methods
Journal title
Journal of computer and system sciences → ACNP
ISSN journal
00220000
Volume
54
Issue
1
Year of publication
1997
Pages
190 - 198
Database
ISI
SICI code
0022-0000(1997)54:1<190:NNWQVD>2.0.ZU;2-C
Abstract
This paper shows that neural networks which use continuous activation functions have VC dimension at least as large as the square of the num ber of weights w. This results settles a long-standing open question, namely whether the well-known O(w log w) bound, known for hard-thresho ld nets, also held for more general sigmoidal nets. Implications for t he number of samples needed for valid generalization are discussed. (C ) 1997 Academic Press.