User: Guest  Login
Document type:
Konferenzbeitrag 
Author(s):
Hao Shen, Martin Gottwald 
Title:
Demystification of Flat Minima and Generalisability of Deep Neural Networks 
Abstract:
Among many unsolved puzzles in theories of Deep Neural Networks (DNNs), generalisability is arguably one of the most puzzling mysteries of DNNs. In this work, we investigates the concept of sharpness/flatness of local minima of the error function, and its relationship to generalisability of DNNs. By defining the sharpness of local min- ima as the largest Eigenvalue of the Hessian, we identify four influencing factors contributing to the sharpness, while three factors are also found...    »
 
Book / Congress title:
International Conference on Machine Learning 
Congress (additional information):
Understanding and Improving Generalization in Deep Learning 
Date of congress:
14 Juni 2019 
Year:
2019 
Quarter:
2. Quartal 
Year / month:
2019-06 
Month:
Jun 
Reviewed:
ja 
Language:
en 
TUM Institution:
Lehrstuhl für Datenverarbeitung