Correlated Parameters to Accurately Measure Uncertainty in Deep Neural Networks

In this article, a novel approach for training deep neural networks using Bayesian techniques is presented. The Bayesian methodology allows for an easy evaluation of model uncertainty and, additionally, is robust to overfitting. These are commonly the two main problems classical, i.e., non-Bayesian...

Ausführliche Beschreibung

Gespeichert in:

Bibliographische Detailangaben
Veröffentlicht in:	IEEE transaction on neural networks and learning systems 2021-03, Vol.32 (3), p.1037-1051
Hauptverfasser:	Posch, Konstantin, Pilz, Juergen
Format:	Artikel
Sprache:	eng
Schlagworte:	Artificial neural networks Bayes methods Bayesian analysis Bayesian statistics Biological neural networks convolutional neural networks (CNNs) Correlation analysis Covariance matrix deep learning Machine learning Measurement uncertainty model uncertainty Neural networks Normal distribution parameter correlations Parameter uncertainty Performance evaluation Predictions Training Uncertainty variational inference
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!

Schreiben Sie den ersten Kommentar!