• 文献检索
  • 文档翻译
  • 深度研究
  • 学术资讯
  • Suppr Zotero 插件Zotero 插件
  • 邀请有礼
  • 套餐&价格
  • 历史记录
应用&插件
Suppr Zotero 插件Zotero 插件浏览器插件Mac 客户端Windows 客户端微信小程序
定价
高级版会员购买积分包购买API积分包
服务
文献检索文档翻译深度研究API 文档MCP 服务
关于我们
关于 Suppr公司介绍联系我们用户协议隐私条款
关注我们

Suppr 超能文献

核心技术专利：CN118964589B侵权必究
粤ICP备2023148730 号-1Suppr @ 2026

文献检索

告别复杂PubMed语法，用中文像聊天一样搜索，搜遍4000万医学文献。AI智能推荐，让科研检索更轻松。

立即免费搜索

文件翻译

保留排版，准确专业，支持PDF/Word/PPT等文件格式，支持 12+语言互译。

免费翻译文档

深度研究

AI帮你快速写综述，25分钟生成高质量综述，智能提取关键信息，辅助科研写作。

立即免费体验

基于图形处理器的贝叶斯套索和多项逻辑回归

Bayesian Lasso and multinomial logistic regression on GPU.

作者信息

Češnovar Rok, Štrumbelj Erik

机构信息

Faculty of computer and information science, University of Ljubljana, Večna pot 113, 1000, Ljubljana, Slovenia.

出版信息

PLoS One. 2017 Jun 28;12(6):e0180343. doi: 10.1371/journal.pone.0180343. eCollection 2017.

DOI:10.1371/journal.pone.0180343
PMID:28658298
原文链接:https://pmc.ncbi.nlm.nih.gov/articles/PMC5489220/
Abstract

We describe an efficient Bayesian parallel GPU implementation of two classic statistical models-the Lasso and multinomial logistic regression. We focus on parallelizing the key components: matrix multiplication, matrix inversion, and sampling from the full conditionals. Our GPU implementations of Bayesian Lasso and multinomial logistic regression achieve 100-fold speedups on mid-level and high-end GPUs. Substantial speedups of 25 fold can also be achieved on older and lower end GPUs. Samplers are implemented in OpenCL and can be used on any type of GPU and other types of computational units, thereby being convenient and advantageous in practice compared to related work.

摘要

我们描述了两种经典统计模型——套索（Lasso）和多项逻辑回归的高效贝叶斯并行GPU实现。我们专注于对关键组件进行并行化处理：矩阵乘法、矩阵求逆以及从完全条件分布中采样。我们的贝叶斯套索和多项逻辑回归的GPU实现，在中级和高端GPU上实现了100倍的加速。在较旧的低端GPU上也能实现25倍的显著加速。采样器是用OpenCL实现的，可以在任何类型的GPU和其他类型的计算单元上使用，因此与相关工作相比，在实际应用中既方便又具有优势。

https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c6ff/5489220/cec4284770d9/pone.0180343.g008.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c6ff/5489220/0a63f86b03f7/pone.0180343.g001.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c6ff/5489220/b14aa5fa2d16/pone.0180343.g002.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c6ff/5489220/b09a230cf5c7/pone.0180343.g003.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c6ff/5489220/402ba8045cce/pone.0180343.g004.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c6ff/5489220/0f736a9846e9/pone.0180343.g005.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c6ff/5489220/2da61c849e1d/pone.0180343.g006.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c6ff/5489220/5d0dfcdb9818/pone.0180343.g007.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c6ff/5489220/cec4284770d9/pone.0180343.g008.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c6ff/5489220/0a63f86b03f7/pone.0180343.g001.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c6ff/5489220/b14aa5fa2d16/pone.0180343.g002.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c6ff/5489220/b09a230cf5c7/pone.0180343.g003.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c6ff/5489220/402ba8045cce/pone.0180343.g004.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c6ff/5489220/0f736a9846e9/pone.0180343.g005.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c6ff/5489220/2da61c849e1d/pone.0180343.g006.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c6ff/5489220/5d0dfcdb9818/pone.0180343.g007.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/c6ff/5489220/cec4284770d9/pone.0180343.g008.jpg

相似文献

1
Bayesian Lasso and multinomial logistic regression on GPU.基于图形处理器的贝叶斯套索和多项逻辑回归
PLoS One. 2017 Jun 28;12(6):e0180343. doi: 10.1371/journal.pone.0180343. eCollection 2017.
2
Fast acceleration of 2D wave propagation simulations using modern computational accelerators.利用现代计算加速器快速加速二维波传播模拟。
PLoS One. 2014 Jan 30;9(1):e86484. doi: 10.1371/journal.pone.0086484. eCollection 2014.
3
GPU MrBayes V3.1: MrBayes on Graphics Processing Units for Protein Sequence Data.GPU MrBayes V3.1：用于蛋白质序列数据的图形处理单元上的 MrBayes。
Mol Biol Evol. 2015 Sep;32(9):2496-7. doi: 10.1093/molbev/msv129. Epub 2015 May 26.
4
A fast forward projection using multithreads for multirays on GPUs in medical image reconstruction.基于 GPU 的医学图像重建中多线程快速前向投影的多射线算法。
Med Phys. 2011 Jul;38(7):4052-65. doi: 10.1118/1.3591994.
5
Heterogeneous computing for epidemiological model fitting and simulation.流行病学模型拟合与模拟的异构计算。
BMC Bioinformatics. 2018 Mar 16;19(1):101. doi: 10.1186/s12859-018-2108-3.
6
Faster Self-Consistent Field (SCF) Calculations on GPU Clusters.在GPU集群上更快的自洽场（SCF）计算
J Chem Theory Comput. 2021 Dec 14;17(12):7486-7503. doi: 10.1021/acs.jctc.1c00720. Epub 2021 Nov 15.
7
Multi-GPU implementation of a VMAT treatment plan optimization algorithm.容积调强放疗（VMAT）治疗计划优化算法的多图形处理器（Multi-GPU）实现
Med Phys. 2015 Jun;42(6):2841-52. doi: 10.1118/1.4919742.
8
A nonvoxel-based dose convolution/superposition algorithm optimized for scalable GPU architectures.一种针对可扩展GPU架构进行优化的基于非体素的剂量卷积/叠加算法。
Med Phys. 2014 Oct;41(10):101711. doi: 10.1118/1.4895822.
9
Medical image segmentation on GPUs--a comprehensive review.基于 GPU 的医学影像分割技术综述。
Med Image Anal. 2015 Feb;20(1):1-18. doi: 10.1016/j.media.2014.10.012. Epub 2014 Dec 2.
10
GPU accelerated generation of digitally reconstructed radiographs for 2-D/3-D image registration.GPU 加速的数字重建射线影像生成，用于 2-D/3-D 图像配准。
IEEE Trans Biomed Eng. 2012 Sep;59(9):2594-603. doi: 10.1109/TBME.2012.2207898. Epub 2012 Jul 11.

引用本文的文献

1
Development and validation of a nomogram prediction model for factors influencing I-refractory Graves' hyperthyroidism.影响碘难治性Graves病甲亢因素的列线图预测模型的开发与验证
Front Endocrinol (Lausanne). 2025 Aug 13;16:1628226. doi: 10.3389/fendo.2025.1628226. eCollection 2025.

本文引用的文献

1
On the utility of graphics cards to perform massively parallel simulation of advanced Monte Carlo methods.论图形卡在执行高级蒙特卡罗方法大规模并行模拟中的效用。
J Comput Graph Stat. 2010 Dec 1;19(4):769-789. doi: 10.1198/jcgs.2010.10039.
2
Understanding GPU Programming for Statistical Computation: Studies in Massively Parallel Massive Mixtures.理解用于统计计算的GPU编程：大规模并行大规模混合研究
J Comput Graph Stat. 2010 Jun 1;19(2):419-438. doi: 10.1198/jcgs.2010.10016.
3
Regularization Paths for Generalized Linear Models via Coordinate Descent.
基于坐标下降法的广义线性模型正则化路径
J Stat Softw. 2010;33(1):1-22.