BriefGPT.xyz
Mar, 2020
随机排名:无标度离散函数的全局优化
StochasticRank: Global Optimization of Scale-Free Discrete Functions
HTML
PDF
Aleksei Ustimenko, Liudmila Prokhorenkova
TL;DR
该研究提出了一个用于直接优化排名度量的高效框架——CatBoost,并介绍了两种重要的技术:随机平滑和基于部分积分的新梯度估计,证明了经典平滑方法可能会引入偏差,并提出了一个通用解决方案进行去偏差,该算法保证全局收敛性并在多个学习排名数据集上优于现有方法。
Abstract
In this paper, we introduce a powerful and efficient framework for the
direct optimization
of
ranking metrics
. The problem is ill-posed due to the discrete structure of the loss, and to deal with that, we introdu
→