Backprop Evolution
Neural and Evolutionary Computing
2018-08-09 v1 Machine Learning
Machine Learning
Abstract
The back-propagation algorithm is the cornerstone of deep learning. Despite its importance, few variations of the algorithm have been attempted. This work presents an approach to discover new variations of the back-propagation equation. We use a domain specific lan- guage to describe update equations as a list of primitive functions. An evolution-based method is used to discover new propagation rules that maximize the generalization per- formance after a few epochs of training. We find several update equations that can train faster with short training times than standard back-propagation, and perform similar as standard back-propagation at convergence.
Cite
@article{arxiv.1808.02822,
title = {Backprop Evolution},
author = {Maximilian Alber and Irwan Bello and Barret Zoph and Pieter-Jan Kindermans and Prajit Ramachandran and Quoc Le},
journal= {arXiv preprint arXiv:1808.02822},
year = {2018}
}