On Performance and Calibration of Natural Gradient Langevin Dynamics

On Performance and Calibration of Natural Gradient Langevin Dynamics

Producing deep neural network (DNN) models with calibrated confidence is essential for applications in many fields, such as medical image analysis, natural language processing, and robotics. Modern neural networks have been reported to be poorly calibrated compared with those from a decade ago. The...

Ful tanımlama

Detaylı Bibliyografya
Asıl Yazarlar: Hanif Amal Robbani, Alhadi Bustamam, Risman Adnan, Shandar Ahmad
Materyal Türü: Makale
Dil:English
Baskı/Yayın Bilgisi: IEEE 2023-01-01
Seri Bilgileri:IEEE Access
Konular:
Natural gradient
second-order optimization
Bayesian deep learning
Langevin dynamics
confidence calibration
predictive uncertainty
Online Erişim:https://ieeexplore.ieee.org/document/10131934/

Internet

https://ieeexplore.ieee.org/document/10131934/

Benzer Materyaller