arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2204.04582math.APcs.LG
收藏

实阶全变差及其在学习方案中损失函数的应用

Real order total variation with applications to the loss functions in learning schemes

Chinese PLA General Hospital(中国人民解放军总医院);Lakehead University(湖首大学)

Pan Liu, Xin Yang Lu, Kunlun He

摘要:

损失函数是现代数据驱动方法的重要组成部分,如双层训练方案和机器学习。本文提出了一个由$r$阶(an)-迷向全变差半范数$TV^r$,$r\in \mathbb{R}^+$组成的损失函数,它是通过Riemann-Liouville(R-L)分数阶导数定义的.我们着重研究了这类损失函数的关键理论性质,如关于函数和导数$r$阶的下半连续性和紧性。

英文摘要:

Loss function are an essential part in modern data-driven approach, such as bi-level training scheme and machine learnings. In this paper we propose a loss function consisting of a $r$-order (an)-isotropic total variation semi-norms $TV^r$, $r\in \mathbb{R}^+$, defined via the Riemann-Liouville (R-L) fractional derivative. We focus on studying key theoretical properties, such as the lower semi-continuity and compactness with respect to both the function and the order of derivative $r$, of such loss functions.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑