English

Deforming the Loss Surface

Computer Vision and Pattern Recognition 2020-09-15 v2 Machine Learning Neural and Evolutionary Computing

Abstract

In deep learning, it is usually assumed that the shape of the loss surface is fixed. Differently, a novel concept of deformation operator is first proposed in this paper to deform the loss surface, thereby improving the optimization. Deformation function, as a type of deformation operator, can improve the generalization performance. Moreover, various deformation functions are designed, and their contributions to the loss surface are further provided. Then, the original stochastic gradient descent optimizer is theoretically proved to be a flat minima filter that owns the talent to filter out the sharp minima. Furthermore, the flatter minima could be obtained by exploiting the proposed deformation functions, which is verified on CIFAR-100, with visualizations of loss landscapes near the critical points obtained by both the original optimizer and optimizer enhanced by deformation functions. The experimental results show that deformation functions do find flatter regions. Moreover, on ImageNet, CIFAR-10, and CIFAR-100, popular convolutional neural networks enhanced by deformation functions are compared with the corresponding original models, where significant improvements are observed on all of the involved models equipped with deformation functions. For example, the top-1 test accuracy of ResNet-20 on CIFAR-100 increases by 1.46%, with insignificant additional computational overhead.

Keywords

Cite

@article{arxiv.2007.12515,
  title  = {Deforming the Loss Surface},
  author = {Liangming Chen and Long Jin and Xiujuan Du and Shuai Li and Mei Liu},
  journal= {arXiv preprint arXiv:2007.12515},
  year   = {2020}
}

Comments

This paper is not perfect yet

R2 v1 2026-06-23T17:22:37.770Z