Can I read Complexity Control by Gradient Descent in Deep Networks on EtoBox?

Complexity Control by Gradient Descent in Deep Networks by Poggio, Tomaso; Liao, Qianli; Banburski, Andrzej is a Biochemistry, Genetics and Molecular Biology article available to read on EtoBox.

What is Complexity Control by Gradient Descent in Deep Networks about?

Overparametrized deep networks predict well, despite the lack of an explicit complexity control during training, such as an explicit regularization term. For exponential-type loss functions, we solve this puzzle by showing an effective regularization effect of gradient descent in terms of the normalized weights that are relevant for classification.

Who reads Complexity Control by Gradient Descent in Deep Networks?

Author

Poggio, Tomaso; Liao, Qianli; Banburski, Andrzej

Publisher

Nature Publishing Group; Springer Science and Business Media LLC; [London]: Nature Publishing Group, 2010-; Society for Mining, Metallurgy and Exploration Inc.; Research Square (ISSN 2041-1723)

Published

2020

Language

EN

Field

Biochemistry, Genetics and Molecular Biology (Life Sciences)