arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2607.03367math.STstat.MLstat.TH
收藏

核Stein差异的极小极大估计:迹与希尔伯特-施密特尺度

Minimax Estimation of Kernel Stein Discrepancy: Trace versus Hilbert-Schmidt Scales

Weierstrass Institute(魏尔斯特拉斯研究所)

Davit Gogolashvili

AI总结:

研究从n个独立观测估计核Stein差异,确定控制极小极大风险的尖锐谱常数,正部平方根U统计量达最优尺度,标准插件V统计量次优。

摘要:

核Stein差异(KSD)用于比较样本与仅通过得分已知的固定目标分布,广泛用于拟合优度检验等。我们研究从n个独立观测估计\(\operatorname{KSD}(P_0,P)\),确定控制极小极大风险的尖锐谱常数,正部平方根U统计量达到极小极大尺度\(\sqrt{\|C_\star\|_{\mathrm{HS}}/n}\),标准插件V统计量在迹尺度\(\sqrt{\operatorname{tr}(C_\star)/n}\),高斯目标下有维度指数差异。

英文摘要:

Kernel Stein Discrepancy (KSD) compares a sample to a fixed target distribution known only through its score, and is widely used for goodness-of-fit testing, sample quality assessment, and approximate inference. We study the estimation of $\operatorname{KSD}(P_0,P)$ from $n$ independent observations and identify the sharp spectral constant governing the minimax risk: it is the Hilbert-Schmidt norm of the Stein covariance operator $C_\star$, giving the minimax scale $\sqrt{\|C_\star\|_{\mathrm{HS}}/n}$. This scale is attained by the positive-part square-root U-statistic, whereas the standard plug-in V-statistic remains at the trace scale $\sqrt{\operatorname{tr}(C_\star)/n}$ and is therefore suboptimal by the fourth root of the effective rank of $C_\star$; for a Gaussian target with a fixed-bandwidth Gaussian kernel this factor is exponential in the dimension.

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。

↑