← 研踪 PaperTrailarXiv:2510.16937 · v2
Prediction-Augmented Trees for Reliable Statistical Inference
首次提交 2025-10-20 01:21
版本更新 2026-11-16 02:51
雷达观测 2026-11-17 17:12
摘要
Machine learning (ML) is increasingly central to scientific discovery, but translating ML predictions into rigorous scientific claims still requires statistical inference on gold-standard labeled data. A setting that has emerged across the sciences provides the analyst with a small labeled sample, a much larger unlabeled sample, and a pre-trained predictor, with the goal of constructing a confidence interval for a quantity of interest, as was formalized in Angelopoulos et al. (2023). In this ML-driven scientific discovery literature, labeled data is typically the binding constraint on precision while computation is comparatively unconstrained, motivating the design question of how additional computation can extract more inferential value from a fixed labeled set, a fixed unlabeled set, and a fixed predictor. To address this question, we propose the Prediction-Augmented Residual Tree (PART) estimator. PART takes the same inputs as the original PPI estimator and replaces its single global rectifier with locally computed rectifiers obtained from an adaptive partition of the feature space. We show that PART outperforms existing methods by producing tighter confidence intervals across real-world datasets from ecology, astronomy, and census reports, among other domains. We then describe and analyze PAQ, an estimator based on numerical quadrature techniques. Under appropriate assumptions on the input data, we show that the variance of PAQ shrinks at a rate of , showing that there are settings where the rate of of existing methods can be provably broken.
作者备注:39 pages, 7 Figures, NeurIPS 2026
观测到的版本
相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。
v2 · 2026-11-16 02:51 · 最新观测版本
Prediction-Augmented Trees for Reliable Statistical Inference
Machine learning (ML) is increasingly central to scientific discovery, but translating ML predictions into rigorous scientific claims still requires statistical inference on gold-standard labeled data. A setting that has emerged across the sciences provides the analyst with a small labeled sample, a much larger unlabeled sample, and a pre-trained predictor, with the goal of constructing a confidence interval for a quantity of interest, as was formalized in Angelopoulos et al. (2023). In this ML-driven scientific discovery literature, labeled data is typically the binding constraint on precision while computation is comparatively unconstrained, motivating the design question of how additional computation can extract more inferential value from a fixed labeled set, a fixed unlabeled set, and a fixed predictor. To address this question, we propose the Prediction-Augmented Residual Tree (PART) estimator. PART takes the same inputs as the original PPI estimator and replaces its single global rectifier with locally computed rectifiers obtained from an adaptive partition of the feature space. We show that PART outperforms existing methods by producing tighter confidence intervals across real-world datasets from ecology, astronomy, and census reports, among other domains. We then describe and analyze PAQ, an estimator based on numerical quadrature techniques. Under appropriate assumptions on the input data, we show that the variance of PAQ shrinks at a rate of , showing that there are settings where the rate of of existing methods can be provably broken.
查看此版本解读 →查看此版本原文 ↗