本文介绍了一种新的诊断工具，用于理解在校准数据被污染时，修剪如何影响一致性预测。研究分析了固定阈值修剪，并非将其视为净化方法，而是作为一种条件化过程，用保留律替换被污染的校准律。研究结果表明，当异常分数能有效分离保留概率而不改变干净总体时，修剪是有益的，并为有限样本证书提供了模板。 AI
影响 引入了一种用于一致性预测的新型诊断方法，有可能在存在嘈杂校准数据的情况下提高模型可靠性。
排序理由 这是一篇发表在arXiv上的研究论文，详细介绍了一种用于一致性预测的新诊断方法。
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
arXiv:2605.06204v1 Announce Type: cross Abstract: Trimming suspicious calibration points is a common response to contamination in conformal prediction. Its effect on clean-target coverage, however, is governed by the retained law induced by trimming, not by the contamination leve…
Trimming suspicious calibration points is a common response to contamination in conformal prediction. Its effect on clean-target coverage, however, is governed by the retained law induced by trimming, not by the contamination level alone. We analyse fixed-threshold trimming as co…