研究人员开发了在统计模型中创建鲁棒置信区间的新方法，特别是针对Efron高斯两组模型。他们的工作表征了在数据污染比例未知时这些区间的最佳长度。研究结果表明，与污染已知的情况相比，区间长度呈多项式下降，而在噪声方差也未知时，性能会进一步下降。 AI
影响 在鲁棒统计方法方面引入了理论进展，可能影响AI模型的评估和不确定性量化。
排序理由 关于统计建模和置信区间的学术论文。
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
新研究详细介绍了Efron高斯两组模型中的自适应鲁棒置信区间
报道来源 [2]
-
Adaptive Robust Confidence Intervals in Efron's Gaussian Two-Groups Model
arXiv:2604.26992v1 Announce Type: cross Abstract: Robust uncertainty quantification is increasingly important in modern data analysis and is often formalized under Huber's model, which allows an $\varepsilon$-fraction of arbitrary corruptions. In many experimental sciences, howev…
-
Adaptive Robust Confidence Intervals in Efron's Gaussian Two-Groups Model
Robust uncertainty quantification is increasingly important in modern data analysis and is often formalized under Huber's model, which allows an $\varepsilon$-fraction of arbitrary corruptions. In many experimental sciences, however, the measurement protocol is well controlled, a…