The Impossibility of Parallelizing Boosting
Machine Learning
2023-08-22 v3 Computational Complexity
Data Structures and Algorithms
Abstract
The aim of boosting is to convert a sequence of weak learners into a strong learner. At their heart, these methods are fully sequential. In this paper, we investigate the possibility of parallelizing boosting. Our main contribution is a strong negative result, implying that significant parallelization of boosting requires an exponential blow-up in the total computing resources needed for training.
Cite
@article{arxiv.2301.09627,
title = {The Impossibility of Parallelizing Boosting},
author = {Amin Karbasi and Kasper Green Larsen},
journal= {arXiv preprint arXiv:2301.09627},
year = {2023}
}