Training Neural Networks with an algorithm for piecewise linear functions
Optimization and Control
2021-05-18 v2
Abstract
We present experiments on training neural networks with an algorithm that was originally designed as a subgradient method, namely the Volume Algorithm. We compare with two of the most frequently used algorithms like Momentum and Adam. We also compare with the newly proposed algorithm COCOB.
Keywords
Cite
@article{arxiv.2105.06503,
title = {Training Neural Networks with an algorithm for piecewise linear functions},
author = {Francisco Barahona and Joao Goncalves},
journal= {arXiv preprint arXiv:2105.06503},
year = {2021}
}