arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.01733stat.MLcs.LG
收藏

ST-BCP:通过非一致性分数转换紧缩后向符合预测的覆盖界

Improving Backward Conformal Prediction via Non-Conformity Score Transformation

Department of Statistics and Data Science(统计与数据科学系);Southern University of Science and Technology(南方科技大学);College of Mathematics and Statistics(数学与统计学院);Chongqing University(重庆大学)

Junxian Liu, Hao Zeng, Hongxin Wei

AI总结:

本文提出ST-BCP方法,通过引入数据依赖的非一致性分数转换来缩小后向符合预测中的覆盖界差距,实验表明该方法有效减少了覆盖差距。

摘要:

符合预测(CP)提供了一个用于不确定性量化统计框架,能够构造具有覆盖保证的预测集。尽管CP会产生不受控的预测集大小,后向符合预测(BCP)通过强制设定预测集大小的上界并估计由此产生的覆盖保证来反转这一范式。然而,BCP框架内马尔可夫不等式引起的松散性导致估计的覆盖界与经验覆盖之间存在显著差距。在本文中,我们提出ST-BCP,一种新颖的方法,引入数据依赖的非一致性分数转换以缩小覆盖差距。具体而言,我们开发了一种可计算的转换并证明其优于基线的恒等转换。广泛的实验展示了我们方法的有效性,在常见基准上将平均覆盖差距从4.20\%降至1.12\%。

英文摘要:

Conformal Prediction (CP) provides a statistical framework for uncertainty quantification that constructs prediction sets with coverage guarantees. While CP yields uncontrolled prediction set sizes, Backward Conformal Prediction (BCP) inverts this paradigm by enforcing a predefined upper bound on set size and estimating the resulting coverage guarantee. However, the looseness induced by Markov's inequality within the BCP framework causes a significant gap between the estimated coverage bound and the empirical coverage. In this work, we introduce ST-BCP, a novel method that introduces a data-dependent transformation of nonconformity scores to narrow the coverage gap. In particular, we develop a computable transformation and prove that it outperforms the baseline identity transformation. Extensive experiments demonstrate the effectiveness of our method, reducing the average coverage gap from 4.20\% to 1.12\% on common benchmarks.

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。

↑