What is the Inductive Bias of Flatness Regularization? A Study of Deep Matrix Factorization Models
Khashayar Gatmiry, Zhiyuan Li, Tengyu Ma, Sashank J. Reddi, Stefanie Jegelka, Ching-Yao Chuang
Abstract
Recent works on over-parameterized neural networks have shown that the stochasticity in optimizers has the implicit regularization effect of minimizing the sharpness of the loss function (in particular, the trace of its Hessian) over the family zero-loss solutions. More explicit forms of flatness regularization also empirically improve the generalization performance. However, it remains unclear why and when flatness regularization leads to better generalization. This work takes the first step towards understanding the inductive bias of the minimum trace of the Hessian solutions in an important setting: learning deep linear networks from linear measurements, also known as \emph{deep matrix factorization}. We show that with the standard Restricted Isometry Property (RIP) on the measurements, minimizing the trace of Hessian is approximately equivalent to minimizing the Schatten 1-norm of the corresponding end-to-end matrix parameters (i.e., the product of all layer matrices), which in turn leads to better generalization.
BibTeX
@inproceedings{
gatmiry2023what,
title={What is the Inductive Bias of Flatness Regularization? A Study of Deep Matrix Factorization Models},
author={Khashayar Gatmiry and Zhiyuan Li and Tengyu Ma and Sashank J. Reddi and Stefanie Jegelka and Ching-Yao Chuang},
booktitle={Thirty-seventh Conference on Neural Information Processing Systems},
year={2023},
url={https://openreview.net/forum?id=2hQ7MBQApp}
}