Training neural networks for and by interpolation
In modern supervised learning, many deep neural networks are able to interpolate the data: the empirical loss can be driven to near zero on all samples simultaneously. In this work, we explicitly exploit this interpolation property for the design of a new optimization algorithm for deep learning, wh...
Prif Awduron: | Berrada, L, Zisserman, A, Kumar, MP |
---|---|
Fformat: | Conference item |
Iaith: | English |
Cyhoeddwyd: |
Journal of Machine Learning Research
2020
|
Eitemau Tebyg
-
Deep Frank-Wolfe for neural network optimization
gan: Berrada, L, et al.
Cyhoeddwyd: (2018) -
Deep Frank-Wolfe for neural network optimization
gan: Berrada, L, et al.
Cyhoeddwyd: (2019) -
Smooth loss functions for deep top-k classification
gan: Berrada, L, et al.
Cyhoeddwyd: (2018) -
Smooth loss functions for deep top-k classification
gan: Berrada, L, et al.
Cyhoeddwyd: (2018) -
Trusting SVM for piecewise linear CNNs
gan: Berrada, L, et al.
Cyhoeddwyd: (2016)