You are here: Home :: Area NEM :: Engineering :: Automation

H. Schmid

A new approach to automatic classification into a large number of classes

Neural networks have been successfully applied to classification tasks with a small (<10) number of classes. Due to the increasing complexity of the resulting nets, classification tasks with a large class number are not tractable with conventional neural nets. This paper describes a new approach to classification into a large number of classes even if only a few classified samples per class are available.

at – Automatisierungstechnik, Oldenbourg Wissenschaftsverlag

Print ISSN: 0178-2312
Volume: 46, 03/1998
Pages: 149

Journal homepage (external site)

Show all available items of this journal

 

Area NEM