BriefGPT.xyz
Mar, 2022
针对GPU的经典机器学习问题优化
Optimization for Classical Machine Learning Problems on the GPU
HTML
PDF
Sören Laue, Mark Blacher, Joachim Giesen
TL;DR
该论文扩展了GENO框架以在GPU上解决约束优化问题，其性能比当前流行的算法（如CVXPY与GPU加速算法）高出数个数量级。该框架使用易于阅读的建模语言描述了约束优化问题。
Abstract
constrained optimization
problems arise frequently in classical
machine learning
. There exist frameworks addressing
constrained optimization
→