arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.00834cs.LGcs.AIstat.ML
收藏

用于准确且稳定分数基密度比估计的最小方差路径原理

A Minimum Variance Path Principle for Accurate and Stable Score-Based Density Ratio Estimation

School of Mathematics, South China University of Technology(华南理工大学数学学院);School of Electronic and Information Engineering, South China University of Technology(华南理工大学电子与信息工程学院);School of Computer Science and Engineering, South China University of Technology(华南理工大学计算机科学与工程学院);Department of Electrical Engineering, Columbia University(哥伦比亚大学电气工程系);Department of Electrical and Computer Engineering, University of Waterloo(滑铁卢大学电气与计算机工程系)

Wei Chen, Jiacheng Li, Shigui Li, Zhiqi Lin, Junmei Yang, John Paisley, Delu Zeng

AI总结:

本文提出最小方差路径原理,通过减少路径方差提升分数基密度比估计的准确性与稳定性,基于Kumaraswamy混合模型实现数据自适应路径学习。

摘要:

基于分数基密度比估计的方法在机器学习中具有强大能力,但面临一个矛盾:理论上路径无关,但实际上路径依赖。我们通过证明实际训练目标与理想的真实目标之间存在一个关键但被忽视的项:分数函数的路径方差。我们提出最小方差路径原理(MVP)以最小化此路径方差。我们的关键贡献是推导出方差的闭式表达式,使优化成为可能。通过使用灵活的Kumaraswamy混合模型参数化路径,我们的方法学习数据自适应、低方差路径,而无需启发式手动选择。这种对完整目标的原理化优化产生了更准确和稳定的估计器,在具有挑战性的基准上建立了新的最先进结果,并提供了一个优化分数基插值的一般框架。我们的代码可在https://github.com/Hoemr/OpenDRE.git中找到。

英文摘要:

Score-based methods are powerful across machine learning, but they face a paradox: theoretically path-independent, yet practically path-dependent. We resolve this by proving that practical training objectives differ from the ideal, ground-truth objective by a crucial, overlooked term: the path variance of the score function. We propose the MVP (**M**imum **V**ariance **P**ath) Principle to minimize this path variance. Our key contribution is deriving a closed-form expression for the variance, making optimization tractable. By parameterizing the path with a flexible Kumaraswamy Mixture Model, our method learns data-adaptive, low-variance paths without heuristic manual selection. This principled optimization of the complete objective yields more accurate and stable estimators, establishing new state-of-the-art results on challenging benchmarks and providing a general framework for optimizing score-based interpolation. Our code can be found in https://github.com/Hoemr/OpenDRE.git.

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。

↑