Deep neural networks with a set of node-wise varying activation functions

In this study, we present deep neural networks with a set of node-wise varying activation functions. The feature-learning abilities of the nodes are affected by the selected activation functions, where the nodes with smaller indices become increasingly more sensitive during training. As a result, th...

Ausführliche Beschreibung

Bibliographische Detailangaben

Veröffentlicht in:	Neural networks 2020-06, Vol.126, p.118-131
Hauptverfasser:	Jang, Jinhyeok, Cho, Hyunjoong, Kim, Jaehong, Lee, Jaeyeon, Yang, Seungjoon
Format:	Artikel
Sprache:	eng
Schlagworte:	Deep network Principal component analysis Pruning Varying activation
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!