研究人员引入了一个名为条件最大均值差异（CMMD）的统一框架来测量条件分布之间的差异。该框架包括各种基于核的度量，例如 CMMD$_0$、CMMD$_1$ 和 CMMD$_2$，并提供了一个通用的 $s$ 级 CMMD。还提出了一种新颖的双重稳健估计器，如果至少一个底层模型被正确指定，该估计器将保持一致。论文通过数值实验表明，CMMD 在统计检验中能有效识别复杂的条件依赖关系。 AI
影响 引入了一种分析条件分布的新统计方法，有可能改进机器学习中的模型评估和测试。
排序理由 介绍新统计框架和估计器的学术论文。
AI 生成摘要 · Google Gemini · 来自 3 个来源。 我们如何撰写摘要 →
Comparing conditional distributions is a fundamental challenge in statistics and machine learning, with applications across a wide range of domains. While proposed methods for measuring discrepancies using kernel embeddings of distributions in a reproducing kernel Hilbert space (…
arXiv:2605.02260v1 Announce Type: new Abstract: Comparing conditional distributions is a fundamental challenge in statistics and machine learning, with applications across a wide range of domains. While proposed methods for measuring discrepancies using kernel embeddings of distr…
Comparing conditional distributions is a fundamental challenge in statistics and machine learning, with applications across a wide range of domains. While proposed methods for measuring discrepancies using kernel embeddings of distributions in a reproducing kernel Hilbert space (…