arXiv：cs.LG· Shahzar Rizvi, David Burt, Vishwak Srinivasan, Renato Berlinghieri, Stefano Del Col, Tamara Broderick·· 7 小时前AI 评分32
面向时空数据的预测驱动推理：跨空间的时间序列估计
Prediction-powered inference for time series across space
AI 导读
研究者提出一种面向时空场景的预测驱动推理方法，在仅有短期带标签时间序列、较长无标签时间序列和不完美预测器的条件下，为每个空间位置提供可靠的未来标签期望点估计与置信区间。该方法针对 PPI 的 i.i.d. 假设在时间依赖下失效、HAC 方法未适配标签填充的问题做出改进，实验显示其优于自然替代方案，已被 NeurIPS 2026 TS-LIMITS Workshop 接收。
正文
Abstract:The following motif is common in spatiotemporal settings: we have a sequence of covariate and label pairs observed for a relatively short, recent time period. We have access to unlabeled covariates over a longer time period. Data is observed over many spatial locations. For instance, crop yield might be observed over a large geographical area for recent years, but weather data (which is informative about crop yield) is available for a much longer period. The goal is to estimate, at each spatial location, the expected label (e.g., crop yield) in the future and provide a valid confidence interval for this value. The observed time period alone is too short for reliable estimates. Imputing missing labels with machine learning can cause substantial bias. Prediction-powered inference (PPI) can correct for this bias, but it relies on an i.i.d. assumption that breaks under our expected temporal dependencies. Heteroskedasticity and autocorrelation consistent (HAC) procedures account for temporal correlation, but have not been adapted to cases where some labels are imputed. We provide reliable point estimates and confidence intervals given: short labeled time series (across spatial locations), a longer unlabeled time series, and an imperfect predictor of labels given covariates. We show our method outperforms natural alternatives.
| Comments: | Accepted to TS-LIMITS Workshop at NeurIPS 2026 |
| Subjects: | Methodology (stat.ME); Machine Learning (cs.LG); Machine Learning (stat.ML) |
| Cite as: | arXiv:2610.08715 [stat.ME] |
| (or arXiv:2610.08715v1 [stat.ME] for this version) | |
| https://doi.org/10.48550/arXiv.2610.08715 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Shahzar Rizvi [view email]
[v1]
Tue, 6 Oct 2026 17:22:39 UTC (35 KB)
来源：arXiv：cs.LG · arxiv.org