This website requires JavaScript.

Iterative regularization in classification via hinge loss diagonal descent

Vassilis ApidopoulosTomaso PoggioLorenzo RosascoSilvia Villa
Dec 2022
摘要
Iterative regularization is a classic idea in regularization theory, that hasrecently become popular in machine learning. On the one hand, it allows todesign efficient algorithms controlling at the same time numerical andstatistical accuracy. On the other hand it allows to shed light on the learningcurves observed while training neural networks. In this paper, we focus oniterative regularization in the context of classification. After contrastingthis setting with that of regression and inverse problems, we develop aniterative regularization approach based on the use of the hinge loss function.More precisely we consider a diagonal approach for a family of algorithms forwhich we prove convergence as well as rates of convergence. Our approachcompares favorably with other alternatives, as confirmed also in numericalsimulations.
展开全部
图表提取

暂无人提供速读十问回答

论文十问由沈向洋博士提出,鼓励大家带着这十个问题去阅读论文,用有用的信息构建认知模型。写出自己的十问回答,还有机会在当前页面展示哦。

Q1论文试图解决什么问题?
Q2这是否是一个新的问题?
Q3这篇文章要验证一个什么科学假设?
0
被引用
笔记
问答