Loss Landscape Sightseeing with Multi-Point Optimization

10/09/2019 ∙ by Ivan Skorokhodov, et al. ∙ 32

We present multi-point optimization: an optimization technique that allows to train several models simultaneously without the need to keep the parameters of each one individually. The proposed method is used for a thorough empirical analysis of the loss landscape of neural networks. By extensive experiments on FashionMNIST and CIFAR10 datasets we demonstrate two things: 1) loss surface is surprisingly diverse and intricate in terms of landscape patterns it contains, and 2) adding batch normalization makes it more smooth. Source code to reproduce all the reported results is available on GitHub: https://github.com/universome/loss-patterns.

READ FULL TEXT
POST COMMENT

Comments

There are no comments yet.

Authors

page 1

page 7

page 8

Code Repositories

loss-patterns

Loss Landscape Sightseeing with Multi-Point Optimization


view repo
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.