Generating simplified computer-implemented neural network

A simplified computer-implemented neural network is generated. The invention relates to techniques for generating a simplified computer-implemented neural network for a device. The method includes receiving a predetermined neural network, the neural network including a plurality of neural network su...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: LIEB DANIEL, NONNENMACHER MICHAEL
Format: Patent
Sprache:chi ; eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:A simplified computer-implemented neural network is generated. The invention relates to techniques for generating a simplified computer-implemented neural network for a device. The method includes receiving a predetermined neural network, the neural network including a plurality of neural network substructures in two or more layers. The method further includes calculating an approximate neural network for the predetermined neural network and calculating a matrix related to the approximate neural network that describes a change in the loss function caused by pruning one or more neural network substructures from the plurality of neural network substructures. The matrix describes a change in the loss function by pruning at least two network substructures simultaneously from a plurality of neural network substructures located within different layers of two or more layers of a predetermined neural network. Finally, the method includes pruning at least one neural network substructure from the plurality of neural ne