BriefGPT.xyz
Jul, 2024
适当损失至少为1/2阶的遗憾
Proper losses regret at least 1/2-order
HTML
PDF
Han Bao, Asuka Takatsu
TL;DR
选择适当的损失函数是机器学习中的一个基本挑战,本文分析了损失函数的性质和与真实概率向量的差异,并研究了基于估计器的预测性能和收敛速度的关系。
Abstract
A fundamental challenge in
machine learning
is the choice of a
loss
as it characterizes our learning task, is minimized in the training phase, and serves as an evaluation criterion for estimators. Proper losses a
→