Max-Diversity Distributed Learning: Theory and Algorithms
Machine Learning
2019-01-21 v2 Machine Learning
Abstract
We study the risk performance of distributed learning for the regularization empirical risk minimization with fast convergence rate, substantially improving the error analysis of the existing divide-and-conquer based distributed learning. An interesting theoretical finding is that the larger the diversity of each local estimate is, the tighter the risk bound is. This theoretical analysis motivates us to devise an effective maxdiversity distributed learning algorithm (MDD). Experimental results show that MDD can outperform the existing divide-andconquer methods but with a bit more time. Theoretical analysis and empirical results demonstrate that our proposed MDD is sound and effective.
Cite
@article{arxiv.1812.07738,
title = {Max-Diversity Distributed Learning: Theory and Algorithms},
author = {Yong Liu and Jian Li and Weiping Wang},
journal= {arXiv preprint arXiv:1812.07738},
year = {2019}
}