Out-of-Core GPU Gradient Boosting
Machine Learning
2020-05-20 v1 Distributed, Parallel, and Cluster Computing
Machine Learning
Abstract
GPU-based algorithms have greatly accelerated many machine learning methods; however, GPU memory is typically smaller than main memory, limiting the size of training data. In this paper, we describe an out-of-core GPU gradient boosting algorithm implemented in the XGBoost library. We show that much larger datasets can fit on a given GPU, without degrading model accuracy or training time. To the best of our knowledge, this is the first out-of-core GPU implementation of gradient boosting. Similar approaches can be applied to other machine learning algorithms
Keywords
Cite
@article{arxiv.2005.09148,
title = {Out-of-Core GPU Gradient Boosting},
author = {Rong Ou},
journal= {arXiv preprint arXiv:2005.09148},
year = {2020}
}
Comments
5 pages, 1 figure